Also by George and Meredith Friedman

The Coming War with Japan

The Future of War

THE
INTELLIGENCE
EDGE
How to Profit in the Information Age

George Friedman
Meredith Friedman
Colin Chapman
John S. Baker, Jr.

C R O W N P U B L I S H E R S , INC.
NEW Y O R K

DEDICATED TO

CONTENTS

W. Edward LeBard
Jonathan Friedman
Maximilian Chapman
Michelle Baker

Netscape Communications Corporation has not authorized, sponsored, or

endorsed, or approved this publication and is not responsible for its tontent.
Netscape and the Netscape Communications Corporate Logos arc trademarks
and trade names of Netscape Communications Corporation. All other product
names and/or logos are trademarks of their respective owners.

Preface

Copyright 1997 by Strategic Foretasting, L.L.C

Introduction: The Knowledge Crisis

All rights reserved. No part of this book may be reproduced or transmitted in

any form or by any means, electronic or mechanical, including photocopying,
recording, or by any information storage and retrieval system, without
permission in writing from the publisher.
Published by Crown Publishers, Int., 201 East 50th Street, New York, New York
10023. Member of the Crown Publishing Group.

VII

From the Agency to the Company:

The Business Intelligence Organization
-'

Space, Time, Money: The Holy Trinity of

Business Intelligence

s.

Putting It Together: The Practice of

Business Intelligence

Random House, Inc. New York, Toronto, London, Sydney, Auckland

http: / /www. randomh ouse .coin/

17

CROWN and colophon are trademarks of Crown Publishers, Inc.

Printed in the United Slates of America
Design by June Bennett-Tantillo

56

I jbrary of Congress Caialoging-in-Publication Data

The intelligence edge : how to profit in the information age / by
George Friedman... [et al.]. 1st e<i.
p.
cm.
Includes bibliographical references and index.
1. Business intelligence. I. Friedman, George.
HD38.7.155 1997
658.47dc21
97-17567
CIP

Find Yourself: Tapping Internal

Resources

84

Mining the Open Source: Passive

Intelligence Gathering

114

Working the System: Semiactive

Intelligence Gathering

142

ISBN 0-60<HK)075-3
10

9 8 7 6 5 4 3 2 1

First Edition

Contents

When You Absolutely, Positively Have

to Know: Active Intelligence Gathering

164

Lucky Guesses: The Art of Analysis

187

Taking Cover: Counter-intelligence

214

And Now a Word from Our Lawyer:

Intelligence and the Law

234

F R E FAC E

Appendix
Index

255
269

This book originated from a series of intercontinental conversations over an extended period of time. The issue was deceptively
simple: Do businesses actually know what they need to know in
order to make decisions? One of the authors, Colin Chapman, is a
longtime business journalist who has been collecting information
on businesses for decades. Two other authors, George and Meredith Friedman, had just emerged from the thickets of collecting
information on the military, during the course of which they had
discovered new sources and techniques that had heretofore been
inaccessible outside the national security community but which
were now pouring forth into die public domain. The question arose
whedier the methods used in tracking information about national
security and the techniques discovered while studying the military
and intelligence communities could be fruitfully applied to the
craft of business intelligence. We were particularly interested in the
new technologies that permitted die intelligence professional to go
beyond espionage to the efficient mining of the public domain.
One of the outcomes of this conversation was the forming of a
business, Strategic Forecasting, L.L.C., which applies these techniques on a daily basis. The other result was this book, which draws
both on our particular perspectives in hunting down information
and analyzing it as well as our experiences operating a business
intelligence service. It was because we were in business dial we real-

VII

viii

Preface

ized that any book on business intelligence that did not include a
word from a counsel was incomplete. Thus, we invited John Baker
to have the final say, making this a wholly realistic work. The purpose of this book is not to provide a comprehensive manual for
business intelligence operations but to create a framework for
thinking about the problem of transforming information into
knowledge. We have also provided some representative anecdotes
to give the reader a sense of the steps involved in this process. All of
the episodes recounted here represent actual cases in which we
have been involved. Since, however, confidentiality is a critical element in our chosen profession, particulars, including but not confined to industry and country, have been occasionally changed.
We would like to thank several people who have helped us in our
collective effort, including Chris Treadaway, Matt Baker, and others
of our staff at Strategic Forecasting who were indispensable in the
research for this book. We are also grateful to a number of members of the staff of the World Economic Forum and its founder and
president, Klaus Schwab. William Bradford Reynolds and Dave Marshall both read significant portions of the manuscript, and we thank
them for their helpful comments. And finally, we thank Bob
Grevemberg, who kept us from taking this too seriously.

THE I N T E L L I G E N C E EDGE

INTRODUCTION
The Knowledge Crisis

Intelligence professionals have regularly regarded their calling as

the world's second oldest profession. They've been modest. Intelligence is actually the oldest profession around. Long before the first
hooker named her price, Gain killed Abel, striking unexpectedly.
From that moment on, humans watched and waited, trying to figure out from where the next threator opportunitywould come.
It is impossible to be human without, in some sense, mastering the
craft of intelligence. From buying a house to selecting a car to taking over a multinational corporation, life is about collecting, analyzing, and acting upon information. Rather than being some
peculiar specialty, intelligence is actually the most ordinary and
common of activities. How odd, then, that business should find
intelligence so exotic!
We associate intelligence with spy satellites, high-speed encryption devices, and, above all, with "Bond.. James Bond." It is hard to
think about garment manufacturers and first-time home buyers as
part of the intelligence game. But, truth be known, government
intelligence is usually about equally unglamorous topics that happen to concern matters of fundamental national interest. And the
same can be said for business and personal intelligenceit's not
glamorous, but it is about as close as you can get to life and death
in everyday existence.
Frequently, we confuse intelligence with espionageillegal

THE

N T E l_l_ I G E N C E

EDGE

activities in the service of the state. But intelligence is much more

than espionage. Intelligence is about information gathering and
analysis. Illegally acquired information may be sexier than the
legally acquired kind, but it is rarely more useful. The most secret,
illegal information improperly analyzed is worthless, and the most
commonly available information properly analyzed can be invaluable. Good intelligence is less concerned with where information
comes from than with making sure that all information required for
decision making is in hand, and that the information has been
properly processed and analyzed.
Information management is at the heart of intelligence. By
information management we don't quite mean what the dataprocessing department means; rather, we mean knowing what to do
with collected information, knowing what is important and what is
not, what can be discarded and what must be preserved, how to
make certain that valuable information is accessible and not lost in
the crowd. This is the core of the intelligence processone that we
can, in fact, learn a great deal about from national intelligence
agencies.
Collection without analysis is like foreplay without orgasm
only more frustrating.
During the Cold War, the Central Intelligence Agency (CIA)
and other national intelligence services were given an enormous
task. They were told that their responsibility was to know everything
going on in the world that could affect the interests of the United
States. Now, since that covered just about everything in the world,
it was inevitable that the CIA would fail. No matter how well they
did their job, sooner or later they would miss something. It was the
very vastness of their mission that caused the CIA and other intelligence-gathering agencies to recognize that their only hope was to
master the information management processto implement a systematic approach to information capture, management, and analysis. The systematic approach actually works better in business than
in government, because business has the luxury of being more
focused.

Introduction

The hard-won lessons of the world's national intelligence services can now be applied to business, and not only large corporations. Any business that needs information about its operating
environmentwhat we call situational awarenesscan profit from
these techniques. What we are proposing is the logical next step in
the great global privatization wave under way since the 1980s: the
privatization of intelligence.
In the simplest sense, everyone in business is an intelligence
operative and analyst. Business is about locating opportunities and
risks, analyzing possible responses and acting on that analysis. Business lives and dies by information collection, management, and
analysis. The issue is not whether you will engage in business intelligence, but rather whether you will do it casually or meticulously,
ad hoc or systematically. Obviously, we think systematic and meticulous intelligence is more profitable than casual, ad hoc intelligence.
You may think otherwise. But doing intelligence is not a matter of
choice. It is what business is all about, and it is getting harder, rather
than easier, to do.

A W O R L D GROWN TOO BIG AND

MOVING TOO FAST

; The world is awash in information. We can sit in New York and

;, know what happened only a few minutes before in a boardroom in
, Hong Kong or a ministry in Brasilia, Government ministries all over
the world are recording, archiving, and allowing access to activities
taking place throughout their societies. Highly skilled and specialized journalists roam the world looking for events to record and
explain. Employees in your own company are accumulating knowledge about things that are vital to decisions being made by your
company's executives right now.
More than ever before in human history, businesses need to
know what is happening on the other side of the world as quickly as
possibleas well as what the mail-room clerk has found out during
Ms lunch break. Opportunities open and close in a matter of hours
and new technologies emerge in weeks and mature in months. New

THE

INTELLIGENCE

EDGE

regulations constantly blindside businesses. All too often, we are

making decisions without enough information. Businesses are suffering from a profound lack of knowledge.
Here is the paradox with which we live on a daily basis:
Although we have more information available today than ever
before, we don't actually know that much more, as a percentage of
the whole, than we used to. We'll make an even more sweeping
assertion: The percentage of information being turned into knowledge today is far less than the percentage a century ago. For all of
our technical sophistication, we are day by day falling behind the
knowledge curve. And it is costing us dearly.
In order to understand this, let's begin by distinguishing
between three things: data, information, and knowledge. Think of
data as a single factfor example, someone just bought a bar of
soap in Chicago; the chairman of the board of General Motors is
sixty-three years old; the Fed Fund rate closed at 5'/a percent on
Tuesday. These are all data. They are true, but, by themselves, they
have only limited meaning.
Now, if we were to say fifty thousand bars of soap were sold in
Chicago yesterday, an increase of 12 percent over the same day last
year, we would have a cluster of factsa set of data. More precisely,
the data would have been manipulated and organized in order to
yield insights that no single piece of data could provide. This clustered and manipulated data we'd call information. So, if we were to
say that the average age of a Fortune 500 board chairman is sixtyone, down from sixty-two a year ago, or that the Fed Fund rate is
declining, we would have taken data and pulled it together so that
it became information. Information is data joined together in
revealing ways so that a person can analyze and understand its
implications.
Knowledge, on the odier hand, exists only inside die heads of
human beings. Entire libraries contain not a bit of knowledge. All
they contain is information, ready to be turned into knowledge
when someone reads, learns, and understands. The crisis of our
time is that the deluge of information that we confront has made it
even more difficult to acquire knowledge. Finding needed information amid the torrent around us is a difficult, sometimes even

/Introduction

impossible, task. Analyzing and understanding that torrent is even

more difficult. Yet without that understanding, neither data nor
information have any value. This surge of valueless information is
the crisis of our time.
Knowledge is what you ire after. Information is the raw
terial you use. Intelligence is what finds and processes
information.
This crisis has arisen because of a single innovation: the comt puter. Think of the computer as nothing more than an extremely
cost-effective recording device. What has happened in business, as
f elsewhere, is that all sorts of events are constantly being monitored
and recorded by computers. Every operation on the factory floor,
every purchase, every sale, every trip, every paycheck, every illness,
every... well, every everything, is recorded. This has made a vast universe of information potentially available to us.
Those of us old enough to remember giant mainframes in airconditioned roomswith less RAM (random-access memory) than
the PC (personal computer) we threw out last yearwill remember
the huge printouts that used to be delivered on little wagons. Huge
data dumps, containing everything we needed to know, yet totally
inaccessible, were placed on our desks. Who could figure out what
those numbers meant? Who could turn the data into information?
The computing industry has spent the last generation turning
all of the data produced throughout the world by computers into
usable information. The PC was one step. By delivering the data to
our desks electronically rather than on paper, it became possible for
us to examine sales reports, production figures, wage ratesor
whatever we were interested inmuch more efficiently. The next
step, of course, was software, designed to turn that data into information. From the simplest word processor and spreadsheet to the
most complex, specialized, and customized software package, the
result of the personal computer on everyone's desktop was
the transformation of available data into information: huge, gigabyte-disk full, terabyte servers full of information. It was a sea of
information, more dian enough to drown in.

THE

NTEUL1GENCE

EDGE

In business today, we have had a massive breakdown in intelligence. To be more precise, the massive surge of information has so
overwhelmed the traditional collection and analysis systems that
businesses use, that most information is lost and wasted. Critically
important information is rarely acquired and even when acquired
rarely turned into knowledge. The information is out there somewhere, but businesses are still making critical and costly errors.
Modern business is facing a massive crisis of knowledge, rooted in
the failure of intelligence. To put it differently, businesses have
become quite sophisticated in turning data into information, but
they are not nearly as good at turning information into knowledge.
This is rooted in the failure of businesses to approach the problem
of knowledge with the same rigorous systems approach they used
in creating first data-processing systems and later informationmanagement systems. They have neglected to create intelligence
systems and to turn information into knowledge.
One solution that companies are turning to, and has been
widely publicized of late, has been the creation of knowledge managers. These people have been given a wide range of titles, from
Chief Knowledge Officer to Chief Learning Officer to Director of
Intellectual Capital and the like. This is becoming a widespread
phenomenon, particularly in large companies that are experiencing the problem first and most intensely because of the size and
complexity of their operations. These companies are not confined
to the computer industry. They include Coca-Cola, Monsanto,
Coopers & Lybrand, General Electric and others throughout the
business spectrum. These people have a common task: to organize,
control, manipulate, and exploit all of the information that has
been created inside the company and turn it into knowledge.
Whatever you call these new people, their mission is the same
as that of the head of any intelligence service: to maximize die efficiency of data collection, collation, and analysis. An intelligence
agency is an organization dedicated to collecting information and
turning it into knowledge. The reason a Central Intelligence
Agency was created, for example, was the conviction, born of experience in World War II, that not only was there tremendous inefficiency in American intelligence gathering during the war, but a

Itroduction

I'great deal of information that was gathered was dissipated because

there was no central collection and analysis point for it. Nor was
there a central distribution point designed to make sure that everyI one who had a need to know would receive the knowledge they
needed.
Today's knowledge managers have as their first mission getting
|" control of knowledge inside their companies. That was the same
task the CIA had on its creation: figuring out what they had and
turning it into usable knowledge. As knowledge managers get
control of their internal systems, they will, inevitably, turn their
attention to systematizing information gathering, analysis, and distributionthe entire knowledge generation cycle. In other words,
the knowledge officer will build an intelligence system for the
company.
It is not that business has been indifferent to intelligence
many large companies have established competitive intelligence
departments. Indeed, the sure sign of this emerging profession was
the creation of the Society of Competitive Intelligence Professionals, to which at least one of the authors belongs. However, we think
competitive intelligence has, in many ways, been too modest. Dedicated to introducing intelligence techniques into the world of business, competitive intelligence has focused on carving out a niche
for itself within the various fiefdorns of the corporate world.
Our view is a bit more radical. From where we sit, carving out
a new division is not only too modest, it is damaging. Business is
already excessively compartmentalized. From our perspective, businesses do not need another class of professionals nearly as much as
the existing professions and specialties within a company need to
add an additional perspective and function to what they are already
doing. From our perspective, intelligence, like the production of
knowledge, must be a universal responsibility within a company
rather than the domain of specialists. That is why our understanding of business intelligence systems is linked to the redefinition of
businesses as systems of knowledge and the symbolic emergence of
the office of Chief Knowledge Officers, Intellectual Asset Managers,
and the like in companies as varied as Coca-Cola and Coopers &
Lybrand.

THE

INTELLIGENCE

EDGE

I Introduction

This is not to say that large businesses don't perhaps require

some specialists in the area. Rather, this argues that intelligence
the knowledge production and management processcannot be
confined to these professionals. Knowledge permeates a company,
flowing into it and being synthesized within it. From capture to dissemination of analysis, the management of die process must involve
the company as a whole. Intelligence systems cannot be created
without creating a culture of intelligence. With that culture, professionals may or may not be necessary. Without that culture, professionals are a waste of space.

INTIMACY AND INTUITION

Now, there may be businesses that do not need to create intelligence systems. A business working with stable technology requiring
only expected, incremental updates; a business working in a circumscribed geographic region, where the laws, regulations, markets, and personalities are well known and changing only slowly;
businesses that are small enough that employees efficiendy share
information; businesses that operate in a stable environment where
the unexpected rarely occurs and time-critical decisions need not
be takensuch businesses, if they exist, would not need intelligence systems.
This is not intended to be ironic. There are, in fact, businesses
where managers are so intimately familiar with their environment
that they sense shifts almost intuitively. A real estate broker, operating in a fairly isolated market, is his own intelligence system. He
knows the few banks in the area personally, as well as the members
of the zoning board. He knows most of his clients personallyand
when he doesn't, he has friends who know them. He knows every
vacant lot, every developer, and knows when a house is likely to
come up for sale, sometimes before die owner does. Someone like
this doesn't need an intelligence system because he is his own intelligence system. His network of relationships, his recollections of
facts about his market and industry, his understanding of the techniques of the realtor substitute superbly for any formal system of
intelligence gathering.

Indeed, if all business was done in this realtor's circumstances,

this book would be unnecessary. But consider how rare such situations are today. Consider all die events that might happen to render our realtor's system of relationships and database of memories
insufficient or even irrelevant: die local banks are gobbled up by
huge bank holding companies; a new plant moves into the community, flooding the area with thousands of new buyers; a federal
agency issues new rules on testing for chemicals on resold properties. All of these things come from outside die neat, stable environment of our realtor, and any of these could decisively change the
way he does business. If he isn't prepared to respond to these
changes, he could be history.
In an ideal business, intelligence systems would be both
unnecessary and irrelevant. Ideally, a manager should be intimately
familiar with all aspects of his working environment, from technology to personalities to regulations, A manager should have such a
degree of intimacy with his environment that his ability to turn
information into knowledge appears to be intuitive. We are all
familiar with successful managers and entrepreneurs who operate
in this way. We are also familiar with people who operate in this way,
have been extremely successful, but who suddenly and unexpectedly seem to lose their touch.
In fact, losing touch is frequently not sudden, even if it's unexpected. It is the result of relying on an intuitive understanding of
the working environment when the needs of the business have
spread beyond the reach of intuition. The surefooted decision making of the earlier days breeds a sense of security and even complacency. The business grows steadily. It is the very success of the
business, its growth, that undermines the intuitive information
gathering and processing into knowledge. As markets grow, personal knowledge of the customers declines. As new product lines
are added, familiarity with technology weakens. And as the number
of employees grows, the all-important casual gossip that was so efficient in sharing information becomes less and less frequent. Many
employees are now strangers.
Suddenly, the entrepreneur is confronting a world where
there is no intimate personal knowledge, and, lulled into complacency by success, he thinks that his old decision-making style will

10

THE

INTELLIGENCE

EDGE

still work. Mistakes begin to pile up. New markets are entered and
sales don't live up to expectations. Nobody within the company
notices new technologies. In the end, the company enters a period
of crisis. The crisis is simple: Intuition works extremely well when
you are intimately familiar with your environment. When success
brings you beyond the reach of intuition, knowledge declines and
information isn't gathered. Increasingly, the business flies blind
until it hits a mountain or runs out of fuel.
You know that informal intelligence systems won't do when
you haven't pliyed golfor pinochlewith most of your
customers.
This is increasingly the entrenched reality of the business
world. The crisis has two roots: geography and technology. Geographythe growth in the size of your company and growth in the
geographical scope of today's businessforces managers to shift
from intuitive to systematic information-gathering and knowledgeproducing techniques. As the network of suppliers and customers
expands, less and less of the market can be knowable. Events completely outside the scope of a manager's intuition can have a devastating effecta fire in a plant that supplies a key component to a
customer that causes a cancellation of orders to your business; a
takeover of a competitor by a large multinational corporation can
cause coronaries at your own shop, even though none of you had
ever heard of the multinational. Forces impinge from far beyond
die realm of expectations. But then, it is rare to be destroyed by the
expected and predictable.
This is one of the critical problems of modern business. Businessmen and -women are practical people. They focus on their own
business and their own work and regard overconcern about the big
world beyond as both extraneous and pointlessthere's not much
you can do about it. This is especially true of entrepreneurs and
those in small businesses. During the 1960s, most auto parts manufacturers focused on producing what General Motors and Ford
wanted. If someone went to them in 1975 and said, "Look, GM is
borrowing money at 13 percent while Japanese companies are bor-

production

11

Rowing at 3 percent," they would probably not have thought it had

(anything to do with them. But they'd be wrong; it had to do with
[efficiency. At 3 percent money, Japanese car manufacturers could
build new plants and facilitiesand put U.S. auto parts manufacturers out of business by the diousands.
You might ask: What could have been done about this? The
ianswer is plentyfrom pressure on congressmen to business plans
focusing on increased diversification to alliances with Japanese auto
inanufacturers. Being small does not mean that you are immune to
global forces. Nor does it mean diat you are helpless. Quite the contrary. The smaller you are, the more likely you are to be subject to
j'1 those forces. And the smaller you are, the more likely you are to be
J able to do something to dodge them.
As companies grow, systems for gathering information in
unknown terrain become critical. As companies become successful,
a more sophisticated definition of information becomes essential.
; Today, geography is one of the driving forces behind the need for
intelligence systems.
The other is technology, or more precisely, the increased rate
of technological change and the growing esotericism of technology.
Early in the industrial revolution, it was possible for one man to
know all about the technology in his industry. He might even know
all about the emerging technologies, since he was responsible for a
lot of them and since the community of researchers was both
extremely small and generally concentrated in a few places. Certainly, inventions could take him by surprise. An inventor in France
might have no idea that someone was doing similar work in the
United States. But there was little that could be done about this.
The information was simply inaccessible.
Today, it is impossible for one person to know everything there
is to know about even a small segment of an industry. Technology
has grown so subtle and nuanced that knowledge has become
tremendously segmented. Moreover, technology has become
increasingly interconnected. Designing a jet engine or a lawn
mower requires knowledge from narrow segments of metallurgy,
petrochemicals, synthetic rubber, and so on. One person cannot
know everything that is happening in these fields that might poten-

12

THE

INTELLIGENCE

EDGE

tially affect his own business. Just as important, technologies

designed to serve the jet-engine industry could, quite easily and
unexpectedly, revolutionize the lawn mower industry. As the texture of technological knowledge becomes finer and more nuanced,
the consequences of technology become less predictable. They also
become more important.
As technology becomes more subtle, the rate of change
increases in two ways. First, small technological advances wind up
having substantial affects on business. Second, the time between
substantial technological change decreases. If we take these phenomena together, it means that the direction from which technological change might come becomes less certain, in the simplest
geographical as well as in a more complex technical sense, while
the frequency of significant change increases. This means that business is both at risk from technological change and that technological change holds open great promise for any business in a position
to take advantage of that change.
We can now define the problem. The geography of business
has expanded beyond the reach of even well-connected and sophisticated businessmen. The best personal network in the world will
not cover all the areas that are of concern to you, nor all of the
social strata you need to know about, nor all the individuals who
might interest you. Modern business is ultimately technologydriven, but the tempo of technological change has increased so
dramatically, and the scope of potentially significant technologies
has expanded so much, that even the most sophisticated businessman or scientist cannot, by his normal daily activities, know everything important that is going on. It has simply become impossible
to master one's environment through casual, day-to-day systems.
You may not be interested in technological change, but
technological change is interested in you.
Finally, because of the computer, much more information is
now available than was the case before. This is an understatement.
The world is pumping out information as a by-product of everything that everyone does, because the introduction of the computer

ntroduction

13

as a facilitator in most human activities has made information a byproduct of everyday life. In a competitive environmentand who
isn't in a competitive environmentthe ability to utilize this infori mation better than your competitors can mean increased sales, new
'products to the market before others, effective joint ventures, avoiding regulatory threats and everything else that constitutes modern
business.
The same type of business, in different locations, generates
dramatically different needs in information. Consider the difference between the needs for information of a painting contractor,
working in a small town in Utah with well-known and slowly evolv* ing technology, who knows all of his customers and competitors,
' and the needs of a painting contractor in a suburb of Washington,
D.C. In Washington, everything is on a larger scale with millions of
potential customers and hundreds of competitors. Whereas the
contractor in Utah knows his competitors and customers personally, the D.C. contractor needs to know how to identify potential
customers, and how to find out what competitors are charging and
paying for help. In a market the size of Washington, D.C., it is not
easy to find out what new products are available, if any new environmental regulations affect contractors, or if there are new,
related businesses.
If you're a painting contractor with two trucks and three workers, you are certainly not going to set up a complex intelligence service, nor are you going to hire an intelligence consultant. But you
may want to think more systematically about the things you do need
to know, and more carefully about how you gather and analyze
information. This may take a little time, a little money, and some
effort in learning a few new tncks, but one or two new contracts,
one or two disasters avoided, and the effort will not only pay for
itself but possibly save your business. So whether you are a large
multinational corporation or a painting contractor with three
employees, you need to turn information into knowledge.
An intelligence system doos not mean an intelligence
department. It means what it says: a system for doing
intelligence.

14

THE

N T E L I- I G E N C E

EDGE

With this, as with anything else, some skills are required. In

most casesparticularly for entrepreneurs operating on a small
and efficient scalethis is a do-it-yourself job. The last thing most
of us need is another professional mouth to feed. But in either case,
and regardless of whatever scale you do it on, the critical issue is
this: Can you make more money using intelligence techniques?
That's what this book is about. How to get a sense of how much
money you're losing through intelligence failures, and how to find
out how to make money through intelligence successes.
In business, time is intelligence's mortal enemy. The intelligence processfrom deciding what information is needed to
understanding its meaningis time consuming. Knowledge a week
too late is the same as ignorance. The problem we face in business
is turning information into knowledge before the decision point
has passed. With the tidal wave of information we face, this is a
daunting task.
Paradoxically, the same technology that created the information crisis also provides us with some of the tools for coping with it.
Computers, information management software, databases, the
Internet, and all the other tools of the modern age are certainly
part of the information-to-knowledge process known as intelligence. These tools, some as sophisticated as remote sensing from
satellites, some as simple as learning to use America Online, are the
front line in the struggle against time and ignorance. They are
indispensable, and we will spend a great deal of this book trying to
teach you how to use them profitably.
But all of the methods and all of the gadgets in the world cannot replace the essence, the core of the intelligence process: honesty, first and foremost, to yourself. The enemy of knowledge is not
only time, it is wishful thinking. It is habit. It is inactivity. It is the
million ways in which someone explains to himself why everything
is just fine, and no exertion is needed. Intelligence works when it
delivers information to a user, who then recognizes that something
has happenedsomething that requires action. Perhaps some previous course was in error and has to be reversed. Perhaps a new possibility is opening up. Whatever the information contains, the user
must have a will to believe, a will to honesty, a willingness to face
potentially painful truths and exert oneself.

Introduction

,5

The mortal enemies of Intelligence are time and wishful

thinking.
In all of our experience in the intelligence process, this is what
has always been most lacking. The will to honesty, the will to believe
what the information shows, both in individuals and organizations,
is the scarcest commodity in the military, in government, and in
business. The ability to see the consequences of actions clearly, even
when that perception runs counter to conventional wisdom,
requires courage and a willingness to be alone. Sometimes it
requires the courage to be wrong, for none of us is guaranteed
omniscience. But the will to believe what the facts reveal, and the
courage to act on those facts, is the foundation of success in all
endeavors.

[FROM THE AGENCY

THE COMPANY
The Business Intelligence Organization

f The Central Intelligence Agency was nicknamed the Company by

$'' its employees. An odd name for an organization that appears to be
'- ]- as far removed from business as possible. After all, the CIA is about
secrets and plots, moles and doubles, the stuff of Le Carre novels.
Companies are about making money. Calling the CIA the Company
is ironic, but in the irony of old agency hands, there is a glimmer of
truth. The CIA in many ways looks like a corporation and acts like
a corporation, a vast global bureaucracy dealing in information. It
is an information conglomerate with one customer: the executive
branch of the U.S. government.
The CIA's interests differ, obviously, from private, profitmaking corporations. Its interest is in national security and its perspective is limitless. That means that it is inherently inefficient from
a financial standpoint. Since threats can come from anywhere, it
must cover everythingand cahnot be limited by cost. This, plus
the fact that, as a government agency, it suffers from the defects of
government agencies everywhere, means that the Company resembles companies in only one sense: It is a vast organization hungry
for information. In feeding this hunger, the CIA has developed real
expertise in gathering and analyzing information. In this sense,
there is much to be learned from the CIA, and from the other great
intelligence agencies around the world like MI-6 in Britain, the old
KGB in the Soviet Union, and Israel's Mossad.
17

IB

HOW

THE

NATIONAL

INTELLIGENCE

EDGE

INTELLIGENCE

AGENCIES A R E O R G A N I Z E D

Intelligence has always been with us, but the modern intelligence
organization really emerged during the last hundred years. The
twentieth century saw intelligence gathering turn from an incidental activity carried out by diplomats, journalists, and travelers in the
course of their other activities, into a systematic, focused process
run by professionals for whom intelligence gathering was the main
activity of their lives. The reasons for this development are rooted
in two phenomena. Internally, the emergence of revolutionary parties meant that the state needed to systematically monitor the opposition in order to prevent or preserve the revolution. Externally, the
age of total war meant that monitoring potential threats to national
security required a systematic and focused approach to intelligence
gathering.
Revolution and total war combined to give rise to an extraordinary phenomenon: the intelligence organization. Unlike anything before in history, the intelligence organization was a vast
bureaucracy created for the sole purpose of gathering and understanding information of interest to the rulers of the nation-state.
Whatever one thinks of the purpose to which any particular
agencyor for that matter, all intelligence agenciesis put, the
sheer scope, the sheer ambition of organizations such as the CIA is
breathtaking. It had as its goal the ability to pursue all information
relevant to a nation's interestvast indeedthe ability to analyze
that information, and the ability to transmit the analysis to decision
makers in a timely and understandable fashion.
What is remarkable about these organizations is how well they
succeeded. We of course hear a great deal about intelligence failures, although many of the failures, such as the Bay of Pigs, were
less intelligence failures than political miscalculations. Yet the intelligence services of modern advanced industrial countries have
provided their political masters, day in and day out, with unprecedented situational awareness. An American president or a German
chancellor knows far more about what is going on inside or outside
his country than ever before in history. Mind you, this doesn't mean

3m the Agency to the Company

19

that they are any better at ruling. A genius with limited information
will rule better than a dunderhead with encyclopedic knowledge.
!1
Nevertheless, all other things being equal, the information pro? vided by intelligence services proves historically to be of enormous value.
There is, therefore, much business can learn from these intelligence organizations. Business and government have this much in
common: both prosper through lots of information properly anas' lyzed and understood. Understanding how intelligence services
organize themselves will not give us a blueprint for business intelligenceGod knows that a local meat processor doesn't need the
overhead of the CIA. It may, however, give businesses a first glimpse
of what is possible through systematic intelligence practices.
Let's begin by considering what is not part of the intelligence
apparatus: security. The guards at the gate, the locks on the doors,
the bodyguards accompanying executives, even the firewalls protecting computers from hackers are part of an organization's security apparatus. They are certainly important and need to be done
but are not, properly speaking, part of intelligence. Guards and
intelligence agents are doing very different jobs with very different
goals. Putting them together is not useful.
In almost all countries, the intelligence agency charged with
collecting foreign intelligencethe nation-state's competitive intelligenceis kept very separate from the agency charged with stopping foreign intelligence agents from spying on them. In the
United States, the two functions are divided between the Federal
Bureau of Investigation (FBI), which does counterintelligence, and
the CIA, which collects and analyzes foreign intelligence. In Britain,
there is MI-5 and Scotland Yard stnd its Special Branch for internal
security, and there is MI-6 for foreign operations. Israel has Shin
Bet, the internal security and intelligence force, and Mossad, which
runs the foreign intelligencegathering operations.
There are two good reasons for keeping intelligence and security separate. The first is the nature of the people involved in each.
Security personnel, from top to bottom, are cops. They may be very
good at protecting property, penetrating criminal organizations,
and identifying security threats, but cops know little, and could care

om the Agency to the Company

even less, about matters of high policy. The FBI, under J. Edgar
Hoover, succeeded in capturing the famous gangster John
Dillinger, but had a great deal of trouble understanding the subtleties of Franklin Roosevelt's unfolding foreign policy. Hoover's
* best solution to the problem of Japanese intentions during World
War II was to arrest every ethnic Japanese in America.

21

GENERIC A R C H I T E C T U R E OF AN
INTELLIGENCE

ORGANIZATION

Cops enforce laws and arrest people. Political leaders, who are
an intelligence agency's customers, are not nearly as interested in
finding the guilty party as in understanding the capabilities of foreign nations, who may be friends one day and enemies the next,
Putting intelligence operations in the hands of the security people
may catch enemy spies, but it won't get you the kind of intelligence
you need to carry out a foreign policy. The two jobs call for very different types of people.
The second reason you want to keep security separate from
intelligence is to keep intelligence honest. By dividing the CIA from
the FBI, MI-6 from MI-5, Mossad from Shin Bet, you provide a
check on the honesty of the intelligence service. An intelligence
service, when it operates properly, can be a political leader's eyes
and earsand sometimes his brains as well. That means it's in the
interest of every enemy and competitor to blind you or make you
stupid. The best way to do this is to subvert your eyes, ears, and
brainsto turn them against you. This is the nightmare of every
intelligence agency. By keeping security and intelligence separate,
you spare your intelligence service from the responsibility of monitoring itself, allowing the chief executive officer (CEO) to sleep
much better at night. It's why you have both a chief financial officer
(CFO) and a comptroller.
Knowledge is power. Power corrupts. Think about it.
So, we begin with two premises. First, intelligence is not about
security. Second, intelligence must be kept separate from counterintelligence. This first sketch of the architecture is important so we
can begin defining precisely what might be useful to a business in
conducting its own intelligence operations. Organizations are not
shaped ehe way they are by accidenttheir form derives from function. If we can sketch the form, we can understand the function.

At the center of every intelligence agency's architecture is its

reason for being: the customer. An intelligence agency's customers
are the people who have to make decisions. The entire purpose of
the intelligence apparatus, its only reason for existence, is to provide the decision maker with the information he needs to operate
the organization he heads. One of the occupational hazards of the
intelligence world is forgetting the customer. Intelligence people
have an inherent tendency to believe that they are at the center of

in*

INTELLIGENCE

EDGE

the universe. It is therefore very important that we pause now. take

a look at that single, lonely box in the upper-left-hand corner of the
chart on page 21 and realize that it is, in reality, about ten thousand
times larger and more important than the complex of boxes below
it. It is even more important than the decision maker. It is the task
of the decision maker to operate that vast and complex box to his
left and the task of the intelligence apparatus to support him in that
task.
Intelligence directors sometimes forget that their prime reason for existing is to support the decision maker in his ultimate role
of responsibility to the government and the nation. Even decision
makers who become too fascinated with the inner workings of intelligence can forget that. A small entrepreneur who becomes overly
fascinated with the intelligence process can quickly forget that intelligence is a means rather than an end. This has to be resisted at
all costs.
There is an even worse condition. Sometimes, intelligence
agencies decide that the decision makers aren't making the right
decisions. People who do intelligence often suffer from paranoia
and megalomania. The paranoia actually helps them do the job,
part of which is to track subtle enemies. Paranoids don't have to be
reminded of this. Megalomania is dangerous because, with the
tremendous amount of information flowing to him, an intelligence
director may develop the illusion that this information constitutes
knowledge. He may believe he knows things about the world that
the decision makers couldn't begin to imagine. There can be a
compulsion to grab die controls away from the political leaders they
serve. Unless an entrepreneur has some pretty interesting personality disorders, this at least is not something likely to come up.
Information isn't knowledge until it is in the hands of the man
who knows what he wants to do and how to do it. In the hands of a
person who has no real sense of the realities of decision making,
information can appear to be knowledge, but it really isn't. Very few
heads of intelligence agencies ever rose to lead governments. Yuri
Andropov, the former head of the KGB, did, but he spent most of
his reign in a coma, so we don't know how he would have done.
George Bush headed the CIA for a while, but we wouldn't call him

f From the Agency to the Company

23

a professional intelligence director. Even so, many would argue that

his presidency is hardly an argument for promoting heads of the
CIA to supreme power. A real head of state can usually slap the
director of intelligence down with little effort. The skills needed to
run a government are different from those needed to operate an
intelligence system.
There is another side to this. Ideally, die intelligence apparatus is not supposed to do anything but gather information and analyze it. How the intelligence is gatheredlegally or illegally, open
source or espionageis immaterial. People who operate things,
from government agencies to corporate departments, inevitably
become advocates for their organizations. That is their job. Unfortunately, they inflate achievements and hide failures. Intelligence
organizations must be indifferent to the outcome of their analysis.
It must make no difference lo them what actions decision makers
take as a result of Uieir information. When intelligence organizations run things, they arc going to be sorely tempted to skew their
reports to protect their operations. As a result, decision makers will
not get clean analysesand may make bad decisions.
We therefore have this paradox. Ideally, an intelligence agency
should do nothing but look and think. But passive looking is sometimes not enough. It is frequently necessary to dig out
informationsometimes a large, costly, and complex operation.
These operationsspyingare the downfall of many intelligence
agencies. Spies need to protect their operations just as everyone
else does. And as we have seen many times, failed intelligence operations are frequently hidden from view by their sponsors.sometimes toppling governments in the process.
Operational intelligence gathering is an ailment that cannot
he avoided. In the end, it can be fatal, but intelligence organizations can and do take steps to limit the damage rogue intelligence
operations can inflict. Distinguishing between operations and
analysis is critical. The analyst's task is to put the different pieces of
information together, to recognize a pattern. For this he needs to
have a deep historical sense, to know where things have been in
order to recognize the patterns he's seeing. He also needs to have
a clear understanding of the intentions, plans, and policies of the

24

THE

INTELLIGENCE

EDGE

decision makers, for without that, he cannot recognize significant

pieces of intelligence. The analyst does not need to know the
source of information, but he does need to have all the information
available. The analyst may not have the most glamorous job, but he
is the pivot of the system. More than anyone else, he can damage
the intelligence organization if he is turned by an opponent. He
knows too much. Aldrich Ames was deadly precisely because he was
an analyst.
Although analysts make up the bulk of an intelligence organization, collection operations is the part that most people associate
with intelligence work. Here, too, in national intelligence services
there is a division between covert operations and open source.
Covert operations, in turn, divide into two parts: what might be
called political warfare, and espionage. These have completely different purposes, but they are pulled together simply because those
who gather intelligence covertly are also frequently in positions to
carry out political operations.
Stricdy speaking, political warfare is not part of the mission of
an intelligence organization. In fact, as an operation that goes
beyond information gathering, it directly contravenes the primary
purpose of intelligence and tends to undermine its utility. All governments have goals and interests, some tfiat cannot be achieved
through direct persuasion or even the direct and open use of force.
The available alternatives range from assassination to blackmailing
or bribing foreign officials, from raising a rebel army to secretly
arming the government. It makes sense that the same people who
are adept at bribing foreign officials to provide secret documents
are in a position to bribe them to stage a coup. Examples of political warfare abound: support of the Contras in Nicaragua by the
CIA; the Mossad's counterterrorism operations in Europe during
the 1970s; the KGB's support for Ethiopia's Marxist coup against
Haile Selassie. Each was a covert operation, carried out secretly, regularly denied by the sponsoring government, even in the face of
overwhelming evidence that it was taking place.
One of the great dangers any intelligence agent faces is falling
in lovewith a country, a source, an operation, an ideology, or a
member of the opposite, or same, sex. Falling in love skews the

From the Agency to the Company

25

agent's judgment by eliminating die paranoia and cynicism that is

the healthy, well-adjusted condition of a field operativethat is,
unless your name is Bond. In case after case, 007 can appear to fall
in love yet not lose his judgment. For most agents, turning him into
a political warrior turns him into a believer, if not in die cause itself
then in the mission. At the very least, he understands that success
or failure will direcdy affect his career. If he predicts it will fail he
will be regarded as a defeatist and therefore will be held responsible for the failure. In any event, after working intimately with his
network of agents to achieve an active end, he becomes, psychologically, part of that network. He ceases to be a critic and loses the distance required for sound judgment.
The Cold War is strewn with cases where intelligence agents
who had been called on to engage in political warfare failed to provide decision makers with critically needed information. The CIA
had been active in the Vietnam War, carrying out political warfare
assignments such as Operation Phoenix, an attempt to destroy the
Communist command structure through assassinations. Having
been made responsible for executing the secret war, the CIA and its
operatives naturally became positive thinkersessential if they
were to do the job they had been given. But they had another job:
to collect intelligence so that the decision makers would have a
clear idea of what was happening in Vietnam. The reality that the
United States was losing the war was incompatible with the positive
mind-set needed to execute the war. As a result, the CIA was instrumental in blowing smoke in Washington's face. National intelligence agencies should not be the guide in this area, as they are
congenitally drawn into covert operations beyond what discipline
and common sense would dictate.
Covert operations do have a clear place: to gather information
that is vital to the national interest but not legally available. In some
countries, this applies to virtually everything. In North Korea, for
example, virtually all information is considered the property of the
state and, therefore, a state secret. It is impossible to collect most
information in North Korea without breaking North Korean laws.
In odier countries, like the United States and Western Europe, far
more information is public than secret. For example, it is quite

THE

I NTE1.LIGBNCE

EDGE

Ifrom the Agency to the Company

legal to rake photographs at American or British airports, but try to

do this in many countries in the Middle East, Asia, or even Italy or
Greece. One cameraman we know was arrested and thrown into
prison in the United Arab Emirates for trying to take some pictures
of an Emirates Airbus taking oft'and landing, even though he was
filming from a public place outside the airport perimeter.
The premise that if something is secret, or illegal to obtain, it's
far more valuable than if it's publicly available, or legally obtained,
is not true, especially when dealing with foreign countries. These
are often merely cultural differences, although they have to be kept
in mind when collecting information outside the United States and
the United Kingdom. The public domain in most advanced industrial democracies is overflowing with valuable information. The
problem is sorting through it all.
lit most advanced, industrial democracies, intelligence work
consists of sorting through the riches available everywhere. In
the rest of the world, most of the effort goes into finding a few
nuggets.
To a national intelligence service it can be a matter of national
interest or national life and death to maintain a balance between
open source intelligence and covert collections, since this controls
what information is emitted into the ptiblic domainand therefore
what is available for capture. In business, however, covert operations are of little value. Quite apart from the physical and legal risks,
the extraordinarily high cost of covert operations makes it a tool
that is not for routine use in business. Any business, for example,
that tried to routinely steal another business's research and development (R&D) instead of running its own R&D shop would very
quickly go broke. It is frankly cheaper to do your own, good
research than to pay the costs involved in stealing it. Obviously,
there are specialized cases where direct corporate espionage might
pay economicallycrime sometimes pays financially if not morally.
But we are drawn always to die lesson of die Soviet Union, which
basically tried to use covert operations as a regular supplement to
its own R&D. Occasionally it worked. In the end, the money spent
on the KGB would have been better spent in research labs.

27

PROM N A T I O N A L INTELLIGENCE
TO BUSINESS INTELLIGENCE

The equivalents of political warfaresabotage, assassination, organized uprisingsand espionage have no place in business intelligence. As a result, most business executives feel that intelligence is
an exotic, but frankly irrelevant, part of their own business operations. This misconception derives from an overemphasis on these
more sexy and glamorous popularizations of the intelligence structure. It fails to recognize that the heart of intelligence, the mining
of open source data, is entirely applicable to businessand just
needs to be put into an appropriate business context. A business
intelligence organization chart would dispense with operational
functions and focus on open source intelligence, containing technical and human intelligence gathering in an even more powerful
form than most national intelligence agencies use.
The business intelligence service has as its customer the executive suitethe CEO, the chief financial officer, head of sales and
marketing, heads of strategic planning, legal, mergers and acquisitions, and so on. Each has different needs, obviously, but they each
have the same responsibility: the prosperity of the company. The
only responsibility that the head of intelligence has is to gather
open source intelligencethat is, information that is generally and
legally available. Open source information is either fully emitted
into the public domain or can be inferred from a variety of forms,
ranging from formal publication to gossip. This is true of all intelligence organizations, but for various reasons, ranging from cost to
the law, it is a fundamental rule for business intelligence.
Organizations constantly* release information: sometimes
planned and controlled releases, such as annual reports or press
releases, and other times they are uncontrolled emissions, such as
leaks to the press by discontented employees or subpoenaed documents. Sometimes emissions are forced during the normal course
of business, such as material submitted to a government agency
while seeking approval for a product or getting a patent. Information can also be released because of a completely legal, deliberate
manipuladon of the company by outsiders, like reporters or canny
business intelligence specialists.

THE

INTELLIGENCE

EDGE

Sometimes companies don't emit information directly, but

leave evidence of information that can be interpreted. Information
sometimes leaves unintended, but unavoidable, hints of existence.
It is much like an animal. Animals can be tracked by following the
unavoidable, incidental consequences of being an animaldroppings, broken branches, warm grass, pieces of fur caught on bushes.
The same is true of companies. The very process of being a company leaves unavoidable traces of information, if not the information itself. These traces usually lead to the information if they are
followed and queried. In order to find out what another company

THE BUSINESS I N T E L L I G E N C E
ORGANIZATION

COMPANY

EXECUTIVE
SUITE

INVESTIGATIONS

INTELLIGENCE
DlflECTOfl

SECURITY

HUMAN

TECHNICAL

29

is doing in a particular R&rD field, you may not be able to get copies
of their research, but you could go to professional conferences
where some of that company's staff researchers might present a
paper on an aspect of their research. When traces left by funding
organizations are tracked and show research grants given to this
company in the same area of research, you have matched enough
traces to safely calculate the information itself. This is not information in die fullest sense, but it is the traces that can be watched,
tracked, and from which the real intentions and capabilities of a
company can be inferred.
These emissions and traces are the quarry of business intelligence. Usually, like emissions from a smokestack, they are created,
emitted, and dissipated into the air, with no one collecting them
and thinking about what they mean. But when these emissions and
traces are tracked down and collected and taken together, they can
be interpreted and understood. Indeed, by being aware of what you
are learning about your competitors, your company can learn to
guard its own emissions and minimize your own traces as you go
about your business.
Locating information is less a matter of sleuthing than of
sorting. We live in a world of too much information rather than
too little.

OPEN SOURCE
COLLECTION

SEMIACTIVE

From the Agency to the Company

ACTIVE

In tracking information, intelligence services divide their

means into two sorts: active and passive. This is a critically important distinction. In order to understand the central difference
between the two, think of how a submarine's sonar works. Sonar is
essentially a listening device that indicates the presence of other
ships by tracking sounds that propagate very efficiently in water.
Sonar operates in two modes: passive and active. Passive sonar is
nothing more than a microphone picking up the sounds that other
ships inevitably make as they pass through the water. Active sonar,
on the other hand, generates its own sounds, which the submarine
emits and which bounce off other objectslarge fish, rock formations, shipsreturning as an echo. Anyone who saw the movie version of Tom Clancy's Hunt far Red October cannot forget the gripping

me

INTELLIGENCE

EDGE

tension of the American and Soviet submarines chasing each other

around the Atlantic, each boat trying to pick up sounds from the
enemy using passive sonar, and sometimes reverting to active sonar
when there was no other choice.

With passive sonar, no one knows that you are out there listening; you have the element of surprise. It also keeps the submarine safe from counterattack. Its weakness is that it relies on the
emissions of other ships. Active sonar, on the other hand, gives you
control over the situation. No matter how quietly the other boat is
running, no matter how carefully they control their emissions,
active sonar bounces off their hull. The cost is that by pinging away,
other submarines will be able to track down the location of your
transmitter.
The difference between active and passive intelligence is the
difference between actively going out and looking for information
and capturing the information that comes to you. It is the difference, for example, between reading a newspaper or walking into a
company, asking to speak to an official and asking him, point-blank,
to answer a question. The first won't get you noticed, and you'll
only learn what has been printed. The second may get you more
information, but it will also get you noticed. In the past thirty years
or so, the growth of technology has created an entirely new sphere
of electronic information that is passive intelligenceinformation
that makes its way to your door by itself. In effect, it captures emissions, uncovers traces, and delivers them to you.
All other things being equal, passive intelligence is much
cheaper than active intelligence. Passive intelligence, like passive
sonar, takes much less energy than active intelligence. It depends
on effective receivers that are always cheaper than effective scouts.
As soon as someone leaves the office to physically locate information, the clock starts running, the price goes up.
For practical purposes, we can divide passive and active intelligence between the kind of intelligence that can be carried out in
your own office and the kind of intelligence that requires you to
leave the office. In reality, we also use a third category we call semiactive, in which you reach out via phone or fax. It is still activethat
is, you can be trackedbut not as active as driving around looking

From the Agency to the Company

31

for things. And it is much less expensive dian that most expensive
and rareactivity, the investigation.
In a way, the distinction between passive and active intelligence can be restated as the distinction between human and technical intelligenceso long as it is understood that these are not
quite identical. Human, or active, intelligence is about spies and
recruiting spies. Technical intelligence gathering, or passive intelligence, concerns the use of technology to peer, without being seen,
into places that would otherwise be invisible.

Businesspeople might assume that technical intelligence has

nothing to do with them. They see themselves as worlds away from
the National Security Agency (NSA). In many cases they're right,
but increasingly some of these techniques are becoming available
and need to be considered. In the end, however, the safest and best
intelligence that can make or break a business is what is called open
source intelligence.

INTELLIGENCE T E C H N I Q U E S
A N D TECHNOLOGY

Technical intelligence focuses on the means of collecting intelligence. There are, generally speaking, two types of technical intelligence gathering that were introduced by national intelligence
agencies: photographic and electronic. Photographic intelligence,
from satellite reconnaissance to airborne reconnaissance to some
guy with a Kodak Instamatic. allows you to see physical objects,
freezing them in time and space, allowing you to examine them at
your leisure and alongside other information. Electronic intelligence allows you to in tercept the electronic signals of the other guy.
This means everything from monitoring his radio talk show to listening in on his telephone calls, mapping his radar, or tapping into
his computer networkall of which the U.S. intelligence community does overseas.
Electronic intelligence is a huge task. Both the United States
and Britain have created separate agencies to carry out electronic
intelligence (the National Security Agency in the United States and

THE

INTELLIGENCE

EDGE

GHQ Cheltenham in the United Kingdom). The United States has

created a new agency, the National Imagery and Mapping Agency
(NIMA), which is part of the intelligence community but will also
be a combat support agency to provide imagery and mapping for
military purposes. It assumed the functions of obtaining and analyzing photographs from satellite imagery. The Central Imagery
Office, the Defense Mapping Agency, and Defense Dissemination
Program Office have all been incorporated in NIMA. Then, of
course, all countries listen to open electronic emissionsradio and
television. Once these are in place, it becomes possible to listen in
on conversations without being physically in the room when the
conversation takes place, and to see things without having someone
on the ground, looking for you. The revolution in how we hear
and how we see dramatically changed the process of intelligence
collection.
It also changed the way in which we manage information. The
electronic transmission of information had two effects. First, it dramatically increased the amount that could be known. Prior to the
electronic transmission of information, conversations between two
people were private, and hence never to be recovered except from
the memories of the participants. Once they were electronically
transmitted, or accessibleto put a nice turn of phrase to electronic eavesdroppingthey could be captured by people who had
never met either of the discussants. The conversations could also be
retained.
It is still widely believed that intelligence gathering is restricted
to governments and government agencies, such as the CIA or the
Internal Revenue Service in the United States or MI-6, the Inland
Revenue, and the Department of Social Security in Britain. Many
do not see much evidence of it in business, and therefore cannot
imagine surveillance being carried out against them.
But use of techniques such as wiretapping is widespread. In
Britain it is estimated that more than a thousand wiretaps a year are
carried out by the police officially. The government's plans to
extend legalized electronic eavesdropping failed when, in 1997,
they were defeated in their bid to give police power to enter anyone's home or office to plant listening devices or copy private

From the Agency to the Company

33

papers if they believed the occupier may have committed a serious

crime. But the debate in the United Kingdom continues. We can
see that the retention of vast amounts of data by private and government intelligence services meant that new methods of storage
and analysis had to be found.
This was the second effect. The agencies survived the deluge
of information by using electronic methods of storage and, most
important, by learning to manage that information. The National
Security Agency, whose antennae listen in on tens of millions of
conversations a day, retains the conversations forever in electronic
storage media such as floppies, hard disks, tapes, and God knows
what else. But this would be useless without a way of sorting through
it all. The real challenge is recognizing which intercepts are worth
listening to, which ones are important, and building up patterns of
information. Storing and sorting data is the heart of technical intelligence. This is the key to business intelligence as well, and one of
the most important things that businesses need to learn from the
professionals.
The United States deemphasized human intelligenceor
HUMINT, as the cognoscenti of intelligence call itduring the
1970s in favor of technical collection techniques, a move that has
been a subject of controversy ever since. The argument in favor of
technical collectioneavesdropping on phone conversations, tapping into computers, using photo reconnaissance satelliteswas
that it was cheaper and more reliable. Human beings are extremely
expensive, and the more valuable the information they provide, the
higher the price. Humans also lie. get confused, or turn out to be
working for the other sideand.worstof all for an agency's budget,
they are in a position to know when they have really important intelligence in hand, and can shake the agency down for it.
On the other hand, humans are also more versatile. They can
go places where there are no electronic emanations, and they can
recognize valuable information directly, even if it isn't on a list of
keywords provided to a computer database. There are obvious
advantages under certain circumstances of using HUMINT. During
World War I, the legendary Mata Hari, a Dutch-born erotic dancer,
was recruited by the Germans as their top agent in France. She

34

THE

INTELLIGENCE

EDGE

danced into the life of a French cabinet minister and became his
mistress, acquiring much information for the Germans before
being brought to court, convicted, and sentenced to death before a
French firing squad. She was able to collect more information
through her personal persuasions than any electronic means could
have done. She was said to have all the ingredients of the glamour
spy: a hedonistic love for life; intelligence; a powerful personality; a
capacity for deception; and an ability to crack or even break men's
hearts. The Mata Hari legend set a style for European intelligence
work that was to last for more than three decades.
Another advantage of using HUMINT is the fact that they can
network. Throughout the forties, fifties, and sixties, several thousand operativesknown as technology collection officers, working
under assignment in various parts of the world undercover as diplomats, trade officials, technical assistants to international bodies like
the United Nations, UNESCO (United Nations Educational, Scientific, and Cultural Organization), the International Labor Office,
the World Health Organization, and UNCTAD (United Nations
Conference on Trade and Development)pulled in, quite legally,
material from scientific and technological journals, used student
exchanges to gather facts, and attended large numbers of conferences. Conferences were of particular benefit because attendance
was upfront and legitimate. Apart from gathering and filing back
official papers, they provided an opportunity for networking and
identifying possible recruits for espionage. International trade
union meetings were especially useful as a venue for establishing
long-term contacts. The Warsaw Pact representatives often found
political sympathizers among the Western representatives present.
The Vietnam War during the 1960s, and its unpopularity among
intellectuals and left-leaning trade unionists, was frequently used as
an ice-breaker, allowing friendships to develop and documents to
be exchanged.
Human beings can do something no technical means of collection can: provide nuance. In 1973, Syria and Egypt built up their
forces on the Golan Heights and along the Suez Canal. Sharing the
American fondness for technical intelligence, Israel knew precisely
what the Arabs were doing. They used photo reconnaissance, signal

From the Agency to the Company

35

intelligence, electronic techniques, and all the rest to map the precise location of all Arab forces. What they did not know, what diey
could not know from technical intelligence, is why the Arabs were
doing it. The Egyptians and Syrians, fully aware of Israel's technical
capabilities, carefully screened their plans. Nothing was transmitted
electronically. All meetings were held in secured rooms where electronic probes were impossible. The Israelis did not have a spy who
had attended the meetings and briefings or slept with a secretary on
a general's staff; they did not have a spy who was in a position to
pick up the gossip in the Presidential Palace in Damascus. Israeli
intelligence knew what they were capable of doing but had to guess
as to the Arabs' intentions and what they were actually going to
doand guessed wrong, with nearly disastrous results. The Arabs
were able to mount a massive surprise attack in October 1973 that
nearly overwhelmed Israeli defenses.
Despite these obvious advantages of HUMINT, it is clear dial
technical intelligence is here to stay, and this is why it is now possible to speak about spinning off some of those capabilities into the
private sector. It is too expensive, time consuming, and in many
cases illegal for businesses to be running strings of human agents
or, for that matter, electronic and signal intelligence projects on
their opponents.
Finding a human being who actually knows valuable secrets
usually costs more than the secret is worth.
But technical intelligence is not only about intercepting messages; it is also about managing the vast array of information that is
now available in electronic form. Photography and radio, with the
ability to preserve images and the ability to transmit those images at
great distances, changed the entire structure of reality. With the
advent of these two technologies, it was possible to see what was
happening as it was happening, to hear what people were saying
without being there in person. More important, perhaps, it was possible to preserve the events in perpetuity. Still, this did not always
clearly tell us what had happened. The mere fact that the Kennedy
assassination had been photographed did not reveal the truth of

40

THB

INTELLIGENCE

EDGE

what happened, as it was limited by the photographer's perspective.

Nevertheless, we have certainly preserved events and transmitted
their occurrence in ways that were utterly unanticipated a century
ago. And today we have capabilities that wouldn't have been anticipated fifteen years ago.
All sorts of events are being recorded today that had never
been recorded before. Every time a sale is made in a store, every
time a product conies off the assembly line, every time a customer
calls, the event is recorded and the record is maintained in perpetuity. One might say that it is the distinctive characteristic of our civilization that nothing is forgotten and everything is preserved.
These techniques for the collection and preservation of information have been fostered by the world's intelligence agencies. These
techniques are now diffusing even more widely.
Intelligence services became adept at taking pictures at great
distances and transmitting those pictures globally. A critical focus of
intelligence services from World War II until the present day has
been the intercepting of electronic transmissions, from radio shows
to telephone conversations, and they have done it very well. But the
very process of electronically recording events by commercial
media has overtaken the intelligence service's own ability to take
photographs. Repeatedly, people at CIA headquarters in Langley,
Virginia, have turned on the television to CNN to find out what is
actually going on. Moreover, the systematic, even compulsive,
recording of events in our everyday lives has found its counterpart in the systematic and compulsive reporting of summaries
of these events. From the commercial media, to the statistical
office at OPEC (Organization of Petroleum Exporting Countries),
extremely accurate and revealing information pours into the public
domain.

MANAGING THE PUBLIC DOMAIN

During the 1970s, the United States orbited a series of satellites

known as Signal Intelligence, or SIGINT, platforms. The purpose of
these satellites was to intercept the stray electronic signals that car-

[From the Agency to the Company

37

ried both publicly broadcast informationradio and television

and private telephone conversations that are carried by microwave
communications as well as by communications satellites. These
SIGINT satellites vacuumed up electronic signals and downloaded
them to the National Security Agency in Laurel, Maryland. Huge
amounts of data were captured, millions of hours' worth of communications daily. All of it was useless, unless the NSA devoted millions of man hours listening to conversations, like Aunt Natasha
complaining about Uncle Boris's drinking, in order to uncover two
minutes of unguarded conversation by a secretary at the Mongolian
Ministry of Defense.
At about the same time that SIGINT satellites were being
launched, the New York Times reached an agreement with the International Typographical Union. Under the agreement, the Times
would be allowed to introduce computers. Previously, type had
been set by hand, using machine-made lead slugs. From now on,
reporters would type their stories into computers, the type would be
set and edited on computer, and dien the electronic code they had
created would be fed directly into the printing press, producing the
day's newspapers.
Suddenly, during the 1970s, information from such diverse
sources as telephone conversations and newspaper stories found
themselves housed in the same, still new, apparatusthe computer.
This had tremendous implications for how we could acquire information, implications that were not at all clear then and are only
fully emerging today.
Telephone conversations and newspaper stories exist in a computer as binary code. A sound, a picture, a sentenceall exist
within a computer as a string of Os and Is. The binary code for a
capital A in ASCII format is 10000001. The binary code for white is
11111111. Any sound can be expressed as code as wellas today's
multimedia computers demonstrate. This was of enormous importance to NSA, which had millions of hours of sound to sort through.
Having captured electronic emissions of analog phone conversations, it was possible to convert them to digitalor binaryform.
Now, all of those conversations could be stored on computer tapes
as vast arrays of binary code. Let's assume that people in the intelli-

THE

INTELLIOENCE

EDGE

gence community were interested in knowing every time a phone

conversation in Nicaragua mentioned the KGB. High-speed computers could scan through the binary code for all conversations captured in Nicaragua, until a match for KGB could be founda
process not nearly as simple as it sounds because of the complexity
caused by variations in voices. Then, someone fluent in Spanish
could listen to the entire conversation that contained the designated term. In effect, the NSA created vast databases of binary code
that were searchable using search programs created by NSA.
The same process took place in the New York Times's press
room. By writing stories on computersin effect, on the first word
processorsevery story appeared in analog form, as a string of letters on paper, and simultaneously, in electronic form, as a string of
binary code. The intention was to save the Times money by simplifying the typesetting process. The unintended consequence, however, was that every word in every edition of the greatest newspaper
in the world now existed as binary code, saved in electronic media.
This meant that the material could be scanned, just the same
as the NSA scanned signal intercepts. Assume that you wanted to
find every mention of the KGB in the New York Times. By loading all
of the data that was already being created to print the newspaper
into databases, you could, with increasing efficiency over time,
locate every bit of binary code that spelled out KGB, call up the
story in which the string was contained, and read it. Suddenly, the
newspaper morgue, where detectives and intelligence agents spent
their youth poring over old newspapers, was replaced by the computer terminal accessing a database of electronically formatted and
preserved newspapers. What had taken days or weeksif it could
be done at allnow could be done in minutes.
Paper can store information for generationsand it can be
used to wrap salamis. Electronic media can store information
far generations and can be searched in secondsbut it
doesn't do salamis. There are trade-offs everywhere.
In less than a decade every newspaper, every magazine, every
company newsletter, literally everything was printed this way. This

| From the Agency to the Company

39

meant that if the material was saved into a database, all of it could
be searched and accessed in minutes. Vast databases such as LEXISNEXIS, Dow Jones, FT Profile, and Dialog sprang up. All they did
was provide containers for the electronic information dial was a byproduct of typesetting technology. They provided large computers,
databases to hold die data and permit queries to be made, and
modems so users could call in. Suddenly, the world of intelligence
was transformed.
The printing press created public domain; the computer
made public domain accessible. The huge volume of information
existing in what constitutes the public domain is contained in newspapers, scholarly journals, corporate annual reports, and television
and radio program transcripts. The problem is not the availability
of the information, it is finding it. Visit the Library of Congress in
Washington, D.C., or any other great archival library and try to find
something. The catalogs are now, thankfully, computerized, rendering the old card catalog system obsolete. They tell you about the
author, title, and the subjects under which the book is classified. But
even computerized catalogs don't tell you anything about what the
book contains.
Any knowledge worth having has some time and cost constraints imposed on it. Having access to material that will take two
years and $100,000 to find is the same as not having access. Intelligence agencies had a choice of mounting an intelligence operation
that had a 70 percent chance of finding the information in two
weeks at a cost of $50,000, or searching through a library that may
or may not have the material, which probably could not be found
hi under a year widi costs that could double or triple the cost of the
intelligence operation. It is no Surprise that the CIA and MI-6 chose
espionage over library research. It was more cost effective.
But the computerization of typesetting created a new and radically different reality. As with most of the computer revolution, its
origins were in the space program. Spacecraft needed computers
onboard to control the complex systems needed to carry out missions. The computers in use during the early 1960s were vast mainframes, many still driven by the vacuum tubes that had powered the
early ENIACS and UNIVACS. Obviously, these could not be taken

40

THE

INTELLIGENCE

EDGE

aboard spacecraft. NASA (National Aeronautics and Space Administration) needed to create an extremely small yet powerful computer. The solution was found in the microchip.
This deliberate creation of computers had another side effect.
During the 1960s, NASA and its military counterparts came to the
conclusion that they were falling behind the knowledge curve, that
they were not accessing all the material needed to keep abreast of
developments in the various scientific and engineering disciplines
needed to create spacecraft. The Advanced Systems Division of the
Lockheed Missile and Space Company, as it was known then, in
Sunnyvale, California (later to be known as Silicon Valley), grappled with this problem starting in the mid-1960s.
The solution they proposed was a radical idea: create a database that would contain data on articles published in all scientific
and engineering journals. Scientists working at Lockheed and in
the space community would be able to call in queries using key
words. Titles and abstracts would be searched and a bibliography of
relevant articles would be generated. There were many problems to
be overcome, ranging from data storage to the time it took early
computers to search databases to user interfaces and so on. But in
due course a database was created. In 1974, on the heels of the computerization of the printing industry, a new service was offered to
the public by Lockheed: Dialog Information Services.
Dialog was a dramatic breakthrough in the way information
was managed. For the first time, it was possible to skim articlesor
at least parts of themin order to locate items that might be of
interest. This meant scientists could locate needed information
more efficiently than ever before, with less risk of missing something critical. For the first time, the limits of the paper world were
overcome electronically. Not insignificantly, the breakthrough
came in a place noted for its secret "black" projects, including technical intelligence projects for the CIA and NSALockheed's Sunnyvale, California, facility. Dialog created a new capability, but it
was, at least initially, focused on a limited literaturescientific.
In the mid-1960s, a company named Data Corporation, based
in Dayton, Ohio, was doing quick turnaround contract work for
Wright-Patterson and Rome Air Force bases. By the late 1960s, it was

I From the Agency to the Company

41

foundering. At the same time, another Dayton business, a paper

company named Mead Corporation, was looking for ways to offset some of the downturns in the paper industry, and decided
high technology was the way to go. Attracted by Data Corporation's Recon Central system for retrieving aerial reconnaissance
photos and a new method for inkjet print, Mead bought out
Data Corp.
Apparendy, Mead was initially unaware that Data Corp. had
contracted with the Ohio Bar Association to set up a full-text, computerized, legal research program. When it became aware of the
contract, Mead hired Arthur D. Little, Inc., a management consulting firm, to evaluate the situation and was told that Data Corp.'s system was superior to any of the competition in computerized legal
research, and that, in fact, there were no adequate systems currently
available in this area. They also reported that there was certainly a
profitable market for an advanced legal database and retrieval system. Upon Little's recommendation, in 1970, Mead's chairman
decided to invest in the project, hired several of the consultants to
manage it, and formed the corporation Mead Data Central. After a
$14 million redesign, Lexis was put on the market in 1973.
Although far surpassing the original plans of Data Corporation and
the Ohio Bar Association, it is important to note that Lexis grew out
of a company that was also involved widi the military. Dayton is the
home of Wright-Patterson Air Force Base, which happens to be the
center of the U.S. Air Force's research and development program,
and therefore one of the most technically sophisticated places on
the face of the globe.
Lexis, and its later competitor Westlaw, revolutionized legal
research. But from the standpoint of business intelligence, the
great breakthrough came in 1979 with the introduction of Lexis's
nonlegal counterpart: Nexis. Nexis stored the complete content of
thousands of newspapers and periodicals on-line daily, and permitted the user to search every word in every article in a matter of
seconds. Almost without notice, a vast universe of published
information became easily and readily available. Once published
and loaded into the database, the information could never be lost.
The entire structure of the relationship of time to knowledge

42

THE

INTELLIGENCE

EDGE

changed, as did the precision with which information could be targeted and retrieved.
Mead Data Central became a rnultibillion-dollar company and
was sold to Reed Elsevier in 1994 for about $1.5 billion. By then, an
entire industry of on-line databases had sprung up. Both Dialog
sold to Knight-Ridder in 1988 for $335 millionand LEXIS-NEXIS
had turned from creating their own databases to reselling databases
created by others. Other resellers had grown up, and database companies were swarming everywhere.
What we have seen in the past generation is both an explosion
in the extent and accessibility of the public domain. The geography
of business intelligence, as well as other types of intelligence, has
shifted dramatically as the focus of the intelligence world has also
been transformed from intelligence collection to management. For
the first time, a systematic and cost-effective approach to business
intelligence has become possible. Now the first search yields
enough for decision makers to make most decisions. Much more
important, the decrease in the time needed for information gathering, and the growing efficiencies of information management,
have increased the time available for understanding the material,
for hunting down the missing pieces, and for analysis, which is, after
all, the entire point of the intelligence process.

SPACE, TIME, MONEY

The Holy Trinity of Business Intelligence

The Book of Genesis tells us the story of Joseph, who has been given
the gift of foretelling the future. This gift gets him into alotof trouble; his brothers sell him to a band of wandering Arabs as a reward
for being right too often. However, the gift also saves his neck, by
making him invaluable to his masters. How valuable was he? In the
first instance of commodity manipulation in recorded history,
Joseph corners the grain market for Pharaoh by interpreting one of
his dreams:
Behold, there come seven years of great plenty throughout all the land of Egypt. And there shall arise after them
seven years of famine Now therefore, let Pharaoh look
out a man discreet and wise and set him over the land of
Egypt. ...And let them gather all the food of these good
years that come and lay up corn under the hand of
Pharaoh for food in the cities and let them keep it. And
the food shall be for a store to the land against the seven
years of famineAnd he gathered up all the food of the
seven years which were in the land of Egypt and laid up
the food in the citiesAnd all countries came unto
Egypt to Joseph to buy com; because the famine was sore
in all the earth.

44

THE

INTELLIGENCE

EDGE

Joseph is subsequently hired by Pharaoh as a commodities broker.

Knowing the future with enough time to react, and having the
resources needed to take advantage of his knowledge, Joseph buys
corn on the spot market and actually takes delivery. Absorbing storage overheads into die low price, he holds until the predicted crop
failures hit. He then turns around and sells in a series of government-to-government cash deals, not only creating tremendous
wealth and shifting the balance of power in the region but also taking some incredible commissions in the process.
Every businessman is a fortune-teller. Unfortunately, few of us
have Joseph's gift from God. Few of us can discern the future by
interpreting dreams, looking at tea leaves, or even from looking at
technical charts of corn futures. What little the rest of us can glean
about the future comes to us from the past. It is not that history
repeats itself in some simpleminded way but that history does in
some sense hold sway over us. After all, most basic structures of the
human condition do remain constant, even as their details change:
agriculture is agriculture, industry is industry, finance is finance,
and so on. These basic structures operate by their own rules, and
while wild variations do occur, life is not, as Macbeth in his madness
would have it, "A tale told by an idiot, full of sound and fury, signifying nothing." If it were, Henry Ford the car magnate, Warren Buffet the well-known financial investor, and the famously wealthy
Rothschilds would never have succeeded in finance and industry.
But they did. Each of them could somehow predict the future when
it came to knowing how to make money.
In order to be a successful fortune-teller, a businessman has to
recognize old and unprecedented patterns, and know how to take
advantage of each. He must be able to identify, almost intuitively,
things that appear to be familiar and therefore may point the way
to what the future might hold. We think about the future through
analogies and metaphors. Things remind us of something; this
product reminds us of an older product; this technology is as new
as some other, successful product; that guy reminds us of a super
guy we once knew. Successful businessmen reason by analogy, but
they also know its limits, dangers, and pitfalls. Analogy alerts us to
possibilities and dangers, but die devil is in die details, and only

Space, Time, Money

45

mastery of the details can provide us with the knowledge we need

to attempt to predict, and influence, the future.
Having mastered the past, businessmen need a constant flow
of information about the present. And they need this information
on two levels. First, they need information on the broad sweep of
things in a constant, unprompted flow from all sources. Second,
they need access to the smallest details on demand, in order to confirm whether their instinctive read was correct or notand to
determine how best to take advantage of the situation.
Successful businessmen are information junkiesand great
historians. You can't spot a trend without the data.

THE G E O G R A P H Y

OF

INFORMATION

Intelligence is always a matter of space and time: Where can information be found and how long will it take to find it? In the particular case of business intelligence, there is a third variable: money.
Every step of the business intelligence process is defined by these
three things. The ultimate luxury would be to find what you need
instandy, on top of your desk, for free. Life rarely makes it so easy.
More often, information takes a long time to find, especially if it's
on the other side of the world, costing so much that it is just not
worth going after. The core problem of business intelligence is to
develop the knowledge and skill to find information quickly and
easily, for as little money as possible.
Businesses and other organizations emit information. Often
the emission is intentional and planned; sometimes it is simply the
uncontrollable by-product of other activities; sometimes it is accidental. Each emission has a trajectory, originating in certain places,
traveling in a certain way, arriving in a certain form. Sometimes it
dissipates and disappears. Sometimes the information congeals, say
into a magazine article, and finds a permanent home in libraries.
Information has a life cycle. It may take a particular form early in its
existence, and over time, change form and location. Information
that is first emitted as gossip may become a confidential internal

46

THE

INTELLIGENCE

EDGE

document somewhere else. It might then be made available in some

informal way, and finally be published in a newspaper and then
made available via an on-line service.
The information capture process is, therefore, extremely
dynamic and complex. We can begin with a basic rule of thumb: the
closer the information, the faster and cheaper it is to obtain. In
planning business intelligence operations, optimizing space is the
basis for optimizing time, which in turn minimizes cost. This means
that a business intelligence system must be built around a geography of knowledge and a method of optimizing the use of timeand
both must be related to price and potential rewards. This is not an
academic exercise. It is a necessary part of focusing a business's
attention on the costs of intelligence.
For many businesses, the apparent cost of doing formal intelligence appears prohibitive, often because there is no clear understanding of the steps involved in controlling costs. If intelligence is
carried out in an uncontrolled fashion, then the question "where
should we look" is always answered with "wherever it takes you"; the
question "how fast do you want it" is answered "as soon as possible."
Answers like these are certain to sour any business on intelligence,
because "do whatever it takes to find it as soon as possible" guarantees that the price will be astronomical. Some projects are worth
this, but not very many.
The geography of information is not quite like a map of New
Jersey. Information exists in a variety of domains, some farther away
and harder to access, some closer and easily accessible. Television,
for example, is part of the electronic domain that comes to you. At
the other end of the scale are competitors' proprietary secrets, that
not only don't come to you but are in a domain so far away that it
is almost impossible to get to them from where you are. Geography
can also be in a more conventional form, such as a particular piece
of information being located in Helsinki. However, assuming that
the information is on a web site in Helsinki, or available via fax, the
actual location is less relevant today than the domain. If you can sit
at your desk and access the information, it is in a very practical
sense "closer" than information physically located in a company
next door.

pace, Time, Money

47

So we must draw a peculiar map to describe the geography of

information, charting the /.ones or domains in which information
resides.

INFORMATION ZONES

Electronically formatted information: This is the first and

most easily accessible layer. Today's databases contain vast
warehouses of readily searchable information, ranging from
obscure industry reports to national newspapers. As time
goes on, this domain will become richer and richer. Its weakness is that most databases contain information from no earlier than about 1980. All of them are accessible with a
personal computer, a modem, and some training. As the
world wide web develops, more and more of this information
will be accessible via the Internet.
Paper formatted information: If information cannot be
found in the electronic domain, the next layer to search is
the paper domain. Paper is an excellent, cost-effective storage medium, and until the advent of the electronic domain,
all published information was formatted on paper. Information on paper, once located, is in many ways the most readily
accessibleeasy to read, to make notes on, and so on. Its
weakness is finding it. While most electronic information
originates in and remains available as paper, not all paper is
available in the electronic domain. In fact, the vast majority
of information produced today is available only on paper. In
the United States, where computers have penetrated most
intensively, electronic information constitutes the largest percentage of all types of currently produced information. In
the rest of the advanced industrial worldEurope, Japan, a
few other countrieselectronic information is being produced, but not at an equivalent rate. Outside the advanced
industrial countriesthis is most of humanitypaper is still
the primary storage medium of information. This makes the
task of business intelligence more difficult, but still doable.

48

THE

INTELLIGENCE

EDGE

ELECTRONICALIY FORMATTED INFORMATION

PAPER FORMATTED INFORMATION
GOSSIP
6RAV ZONE

PROPRIETARY
AND

SECRET

Gossip: As we know in our own businesses, a tremendous

amount of information is known and passed about by word
of mouth, but much of it is never actually recorded and preserved. For want of a better term, we'll call this domain "gossip." It is hard to quantify how much information is
contained in this domain. There is a certain type of information here that is extremely valuable, which might be called
early warnings and unofficial truths. Information is frequently known before it is published, so monitoring this
domain can be complex and demanding. Techniques for
acquiring gossip range from the schmooze to debriefing the
sales force.
Gray zone: Between gossip and the world of secret and proprietary informationas much a legal concept as an operational onethere lies a vital, intermediate zone that we call
the gray zone. This is an area filled with information that has
been formally recorded and is technically available but cannot be accessed by formal means. In other words, it's there
for the asking, but no one knows that it's there, and therefore no one knows it can be requested. This is a zone that
can best be accessed from one of the outer three zones: a

Space, Time, Money

49

casual mention in an article, gossip over drinks, a blind

phone callany of which can reveal the existence of information of the exact type that you need but would have never
found without working at it.
Typical information in the gray zone might be a government report, such as a U.S. Department of Energy internal
study of the future of the natural gas industry in Tunisia. The
study hasn't been published, and only five or six copies exist.
However, it is not secret, and the author, who has never
received any recognition by anyone at all, would probably be
delighted to find that someone cares enough to request a
copy. The problem is finding out that it exists in the first
place. Similar situations occur in the private sector, where
everything from scientific studies to marketing reports is
available, sometimes for free, sometimes for a lot of money
(and not necessarily with any correlation between cost and
value), but only to those who know, or find out, they exist.
Proprietary and secret: This is where open source intelligence ends and espionage begins. Every organization, government, business, and family has its secrets. Sometimes
these are extremely valuable. Sometimes secrecy is used to
make the person who knows the secret feel better about himself. However, the mere fact that something is secret doesn't
mean that it is valuable and an enormous amount of time,
money and risk can be allocated to getting hold of a secret
document only to find out that it contains nothing that could
not have been gotten out of the Wall Street Journal. Moreover,
it is frequently possible toMnfer what is going on deep inside
a company by studying the things that have emanated out
into the public domain, and figuring out the thinking that
went into producing that, and other, information.
The five domains of knowledge are each connected. A search
of an electronic database reveals the existence of a published, paper
document. Reading that document gives you the name of a key
researcher. You call and chat, get together for drinks, and he shares
gossip about another researcher who has developed some interesting studies. You call the other researcher and he agrees to send you

50

THE

INTELLIGENCE

EDGE

a copy of some of his work. In reading his reports, and some other
material you may have, you suddenly realize in what area a competitor of yours is doing research and development. The life cycle
of every bit of information is unique. Some exist in different
domains simultaneously, while some exist in only one. Some are
permanent, others ephemeral, passing away with surprising rapidity. But all exist in one of these domains, and mastering them is vital
to doing business intelligence. And of all the skills of mastery, none
is more important than speed, since time really is money in business
intelligence.

THE

KNOWLEDGE C U R V E

The purpose of business intelligence is to make money. It does this

by providing information about some aspect of the business environment, supplying the business with four advantages. First, it must
provide significant information, information that is not otherwise
known. It must produce something that surprises you.
Second, it must be actionable informationproviding the
business with opportunities for action, information that allows a
business to dodge a bullet. This kind of information costs money.
This question must always be asked: If I knew a particular set of
facts, would I be able to exploit the situation? Would it matter?
Third, the information must be timely. It must be delivered
prior to the event, and it must be provided to the decision maker
with enough lead time for him to respond. For example, a telegram
ordering a war alert was delivered to Admiral Husband E. Kimmel
at Pearl Harbor three hours after the Japanese launched their
attack. The information was certainly significant, but it was not
delivered inside a time frame that was actionablemaking it not
only useless but downright embarrassing. The precise definition of
timely depends on the context. Sometimes it simply means knowing
before the event. Sometimes it means knowing before competitors
know about it. Sometimes it means knowing information so that
other information can be understood, and so on. But this much is
clear: Intelligence can be a wasting asset, and its acquisition must be
measured by time.

[Space, Time, Money

51

Fourth, the information must provide a value greater than its

cost. If it costs $100,000 to get information that is significant, actionable, and timely but with an added value of only $50,000, it will not
take long for intelligence "coups" of this sort to take a company
down.
All intelligence operations must be put to a rigorous test
that measures these four variables: Would the knowledge gained
in this operation be significant, actionable, timely, and cost effective? As with many things in business, it is easy to imagine that
these things would be easily quantifiable, but in fact they are not. It
is difficult to know the precise value of information. A particular
piece of information, by itself, may provide little or no value. Its
value is in the context of pieces of information the existence of
which is as yet unknown. For example, knowing that two people had
lunch yesterday may have no intrinsic value beyond the pleasure of
gossip. But taken together widi a piece of unanticipated information from tomorrow's newspaper, which tells you that one of the
lunch's participants filed for bankruptcy, and knowing that the
other luncher was a banker, might tell you something of the financial position of that company, which may or may not help you,
depending on your position. So, information can have unanticipated and serendipitous value. But most information does not, and
collecting information on the off-chance that it will become valuable is something that will quickly undermine both the bottom line
and your sanity. So, while it may not be possible to rigorously quantify the information, some careful valuation prior to acquiring it is
necessary.
It is important to understand what you are buying when you
create an intelligence system.'But in most cases, the information
you are buying will eventually come to you anyway, as you read the
newspapers, talk to colleagues, or just live a normal life. It's kind of
like standing in Times Square, as the old saying goes: Everyone you
know will eventually pass by. An intelligence system is designed to
get you information when you need it.
Libraries ire loaded with information, most of it free.
Intelligence costs money, because it's about time rather than
information.

3Z

THE

INTELLIGENCE

EDGE

Information is cheap if you don't care how long it takes to

get it. But time is expensive, and each successive unit of time
purchased tends to cost more. In terms of our geography of
information zones, the deeper you go, the more expensive it
gets. Now, those zones pass from published to nonpublished
areas of knowledge. Accessing published information is obviously
cheaper than accessing unpublished or nonpublishable information. This is so if for no other reason than that published
information was intended to be accessed, and was provided for
that purpose, while gossip, the gray area, and proprietary information don't give themselves up nearly so readily. Therefore, the price of acquiring additional warning time rises dramatically as we push the access point back to prepublication
zones.
In thinking about the relationship between information, time,
and money, we can begin to see clearly the questions involved in the
intelligence process. The most important question is: How much
information is enough? A knowledge curve emerges:
THE KNOWLEDGE CURVE

TOTAL EXPENDITURES

53

The time/money/information curve is an S-shaped structure,

; representing early and late inefficiencies in information gathering.
The curve intersects with two other lines. The horizontal dotted
line represents the minimum amount of time that needs to be purchased in order to permit decision makers to make a decision. That
defines the zones that have to be penetrated and the techniques
that need to be used. The solid vertical line represents the maximum value of reacting to the informationreflecting both the benefits of success and the costs of failure.
The knowledge curve is that segment of the time/money/
information curve that is above the reaction line but to the left of
the maximum expenditure line. It represents both the minimum
amount of time that has to be securedthe last point at which the
information can be utilizedand the maximum amount of money
that can be spent. It is the area where intelligence operations can
turn information into knowledge, and where knowledge can precipitate action.
Every acquisition of information, therefore, has a unique
curve, defined by the time that is being purchased and the maximum amount of money that can be spent acquiring that knowledge. If acquisition falls outside of this area, it is impossible. It is
within this area that the information gatherer has choices to make,
including the critical one: Where does he want to be on the knowledge curve?
The core decision in every intelligence project is: How much
time do you want to buywhere do you want to be on the
knowledge curve?

MAXIMUM
RETURN

TIME TO aerial POINT

ace. Time. Money

MINIMUM
REACTION TIME

That, of course, depends on the mission. He can save money,

buy less time, and take longer to acquire the information. He can
spend more money and buy more time. How much time and how
much money are both described in the curve, which, in turn, is
defined by the idiosyncracies of the particular piece of information,
as well as the geography of knowledgewhere within the five zones
the knowledge is to be found. This is the pointthe intersection of

54

THE

INTELLIGENCE

EDGE

space, time, and moneywhere the intelligence specialist defines

his project.

CONCLUSION: B U Y I N G TIME,
CONTROLLING SPACE

The knowledge curve is, obviously, highly stylized. Its primary purpose is to call attention to the peculiarity of intelligence. It is less
about information than about time. In the end, almost anything
important becomes known. But the end can come after a business
has buckled from a catastrophe. The purpose of a national intelligence agency is to provide intelligence that can be acted onwhat
is called actionable intelligence. In a world where an intelligence
failure can cause almost limitless and catastrophic damage, cost is
no object. So, if you were looking at the CIA's knowledge curve, it
would lack the vertical line. It is no accident, therefore, that the
CIA's budget runs into the tens of billions. The earliest point on the
curve must be the access point.
For businesses, where the primary mission is making money,
the vertical line is critical. It represents the limits of effort. It is obviously easier to assert these limits than to calculate them ahead of
time. In later chapters we will discuss techniques of costing intelligence operations. Estimating reaction time is no easy matter either.
But the knowledge curve is the point on which business intelligence
must be fixated. The goal must be to spend enough to get hold of
information while it is still useful without spending so much that
the project's value is destroyed.
When looking at the knowledge curve, you can begin to
understand more clearly why business intelligence is possible today
in a way that wasn't possible even ten years ago. Imagine that Zone
1electronic informationdidn't exist. The steep slope after startup would not exist. Information accumulated in this zone would
have to be back-loaded into other zonespaper, gossip, gray, and
proprietary. Without the efficiencies of the electronic medium, the
curve would slump down and to the right. In most cases, the entire
curve would wind up below the reaction line and to the right of the

Space, Time. Money

55

maximum return line. In other words, business intelligence would

be impossible.
The efficiencies of the electronic zone, which allows a tremendous amount of information to be accumulated very quickly and
relatively cheaply, make it possible to use the other zones more efficiently, targeting missing, needed information rather than using
these zones to accumulate baseline information. In short, the system works only because of the powerful advent of Zone 1.
As we said earlier, making money in business depends, to a
great extent, on fortune-telling. So far, we have not discussed the
reading of tea leaves or even, for that matter, the gathering of tea
leaves. We have discussed the need for tea leaves and have marveled
at the new varieties that have become available. In other words, we
have been setting the stage for the real work.
That begins now. The real work is to master the techniques
and skills needed to mine the five zones efficiently so the intelligence process can hit the knowledge curve. But as we get down to
the nitty-gritty, it is important to bear in mind the core of intelligence: time. Time to react. Time to think. Time to foretell the
future. Mastering intelligence is mastering time, in a world where
time is money.

utting It Together

PUTTING IT TOGETHER
The Practice of Business Intelligence

The foundations of intelligence are discipline and an imposed honesty. An intelligence collector or analyst who lies, particularly to
himself, may as well not begin the process. Two of the deadliest lies
consist of hope and confidence. This is an odd thing to say, since a
positive attitude has always been regarded as one of the keys to success in business. But, in intelligence, as opposed to business in general, a positive attitude is dangerous. Optimism tends to skew your
judgment. The businessman or intelligence specialist doing the
intelligence work always wants to report positive, hopeful things. No
one wants to walk into a meeting and announce that he has just discovered that a new technology by a competing company has
destroyed a project his own company has been working on for two
years. Yet this is precisely the sort of news the intelligence specialist
all too often must report.
As a member of the team, you begin with shared enthusiasm
for a project. It is the intelligence specialist's job to lose his enthusiasm first, to doubtto be right when no one wants him to be
right, even when the careers of others are at stake. Pessimism is a
much harder discipline to maintain than optimism, especially in a
business environment. Yet few things have helped companies more
than doses of truthful pessimism at strategic moments.
Another mortal sin of the intelligence specialist is self-confidence, which can be akin to self-destructiveness. The ultimate discipline is to question your own doubts; to constantly ask if your facts
56

57

are sufficient, if there are any facts available that you haven't collected that would blow your analysis apart, if your analysis itself is
correct, and if your recommendations are on target. If you are
going with the consensus, you must question if you are doing it for
the sake of being one of the crowd. If you are going against the consensus, you must wonder whether you are taking that stand to be
ornery or to show off. And having taken a position, you must wonder continually whether it was the right oneand continue to collect information and evaluate it even after filing your report.
Ultimately, like an auditor about to file a troubling report, the
intelligence specialist has to have courage. Most of us, under pressure, are unpredictable. Accountants range from being Rambo to
being wimpsand sometimes we won't know which until the chips
are down. But when the chips are down, the accountant can fall
back on an accounting process that has enough precision in it to be
looked at and evaluated dispassionatelyit squeezes out hope and
misplaced self-confidence. The process even substitutes for
courage. Right or wrong, the process gives the practitioner something to hold on to. It disciplines, guides, and ultimately controls
him. Therefore, to be successful, we need a process to encourage
the ability to be appropriately negative.
Unlike accounting practices that have evolved over centuries,
business intelligence, as a formal discipline, is in its infancy. Therefore, the world of business intelligence will not have a body of rules
like those that support a lawyer or accountant. At this point, however, we can provide a fairly clearly denned, step-by-step approach
that can guide the intelligence processand that can serve as a
reality check. The intelligence process is not intended to be a straitjacket. Nothing is ever as neat as a flowchart would lead you to
believe. But it is intended to help clarify the processand to move
us from the theory to the practice of business intelligence.
This chart could have been complicated ad infinitum, and by
the time we finish, it will be more complicated. For now, we provide
a fairly simple schematic of the process (see page 58), and describe
a case we worked on showing how it fits into this process: the case
of the magnetic water treatment deviceour first example of a
complete business intelligence project.

58

THE

INTELLIGENCE

EDGE

THE INTELLIGENCE PROCESS

BEFINE
THE
MISSION

IDENTIFY THE
KNOWLEDGE
BASE

ACQUIRE
INFORMATION

SELECT
THE
METHOD

OVEHLAV ON
KNOWLEDGE
BASE

RET ASK
THE
SYSTEM

ARCHIVE
AND
DATABASE

THE CASE OF THE

MAGNETIC WATER CONDITIONER

A client came to us who had put money into a company that manufactured and sold magnetic water treatment systems designed to
reduce the buildup of calcium scale. The company was having problems and he wanted to know if there was any point in putting more
money into it or if he should just cut his losses and run. He understood that this was an intelligence problem, because he felt that
he just didn't know enough about the deal to make an intelligent
judgment.

i'putting It Together

59

In preparing for our first meeting with him, we did a very

quick check on the idea of water magnetism. In doing such checks,
we rely heavily on an electronic database service we've already mentioned: LEXIS-NEXIS. NEXIS, the side that covers news and other
sources extensively, goes back for about fifteen years with decreasing coverage. As a Zone 1 resource, it is fast and efficient. By
inputting the term magnetic water treatment, we quickly got back a
group of stories that would allow us to prepare for the meeting.
LEXIS-NEXIS can appear quite expensive. Running down a single,
simple story can cost hundreds or even thousands of dollars, so it is
not for the casual user. But spending a few hundred dollars familiarizing ourselves with a topic prior to meeting with a client is
a small investment. If it is a client worth going afteror an issue
of fundamental importance to your businessit is worth the financial investment, and of course, the investment of staff and your
own time.
Unfortunately for our client, the news wasn't good. We found
that magnetic water treatment had been around for a while and was
asserted, by its supporters, to be a way to reduce calcium scaling in
water. Scaling is a serious problem in anything that uses water over
an extended period of time, such as boilers in industrial plants.
Scale buildup leads to rust in the ballast holds of ships or automobile radiators. The usual treatments are various chemicals, draining
and scraping the container, or letting the buildup continue until
the vessel is changed.
The theory was that by treating water with magnetism, it somehow reduced scaling. There was, however, a problem. The general
consensus of the scientific community was that this was junk science, that it straight out didn't work, and that diose who claimed it
did were frauds. But we found a lot of anecdotal evidence, some
from people with good reputations, claiming that, despite the consensus, magnetic water treatment did indeed work.
Our client apparently believed in his desalination system. After
all, he had put good money into it. Since we had no basis on which
to form any opinion, and really didn't have expertise in this area,
we neither believed in it nor disbelieved. By trade we are pessimistsmost promising technologies dial no big company wants

60

THE

I N T E 1_ l_ I G E N C E

EDGE

to touch are usually not worth touching. At first glance this seemed
a case where an investor, not himself a specialist in the field, had
taken a plunge and lost. It happens.
We went to the meeting in a suitable frame of mindskeptical. More than anything, this was a last-gasp meeting: Was there anything to be done before he pulled the plug? In such cases there is a
temptation to keep up the investor's hope; after all, hopeful
investors write checks to intelligence companies while hopeless
ones do not. This is a temptation we try not to indulge. To skew an
analysis in order to make money would not only sully our reputation, and the industry's, but in the long run will cost more money
than it earns. But let us not fool ourselves that these thoughts do
not cross our minds; the important thing is to realize their danger.
Always open an intelligence operation in a neutral to negative
frame of mind. It's marketing's job to lead the cheers, net
yeurs.
Our client was enthusiastic enough for both of us. This was fortunate, as we only had to sell our skills and not the value of the project. The conversation turned to the core issue: validation. This
product had the reputation of being fraudulent partiy because of
the involvement of a string of scam artists. To determine whether or
not to put more money into it, or write it off to experience, our
client needed to go back to reexamine the foundations of the business. That was our job: to gather the information he needed to
make his decision and to support his decision making with our
analysis. His initial investment hadn't been enormous. But if he
could produce and sell this product, the calcium buildup problem
he'd be solving would be sufficiently widespread to provide impressive multiplesand lots of money. And so we began our process.
Step 1: Defining the Mission
It was time to get down to work, and all work begins with clearly
defining what exactly we were going to be doing about the magnetic treatment of water. Most intelligence operations suffer from

Putting It Together

61

vagueness and vastness. "Tell me everything" guarantees two things:

The mission will fail, and accounting will go ballistic. No one can
find out "everything," and even trying and failing will bust the budcet. We were once asked by a company to find out everything about
another companyand to do it fast. We tried to explain that what
they were asking for was not only impossible but the bill was likely
to be astronomical. The company they wanted to learn about operated in Russia, England, and the United States. They waved our
warning away and told us to proceed. We probably didn't find out
everythingwhat does everything look like and when do you know
you've found it?but we found out an awful lot. When the bill
arrived, they were... well, it wasn't a pretty sight.
Any intelligence operation must be finite. If it doesn't have a
clear and limited goal, it will fail, literally by definition.
The first step is to honesdy and clearly define the mission:
What do you need to know and why do you need to know it? If you
can't define this with some precision, it usually means you have not
thought tfirough the entire project clearly enough. Indeed, a full
accounting of what you need to know in order to proceed is a general business discipline, and that ought to go hand in hand with
financial accounting. Unfortunately, accounting deals with a relatively simple variablemoneywhile information is complex and
subtle. This increases the need for information accounting, and
controlling it becomes all die more vital in order to focus business
operations in general and intelligence operations in particular.
If you are asked to provide information on another company,
the first bit of information you'need is not about that company but
about your own client. The reason your client is interested in that
company defines the mission. If the CEO is doing his own intelligence work this should not be a problem. He should know what he
is after and where his interests lie. But it is a critical issue in larger
businesses where the intelligence function might be carried out by
someone other than the CEO. All too often, this guy is kept out of
the loop. He gets requests for information without being told why
anyone is interested in it. As a result, he often doesn't know what

62

THE

INTELLIGENCE

EDGE

he's really after. Like the poorly trained dog, he keeps fetching
everything but the master's slipper. In the end, he is held responsible for the intelligence failures when in fact the real responsibility
lies with the CEO for not communicating honestly with his intelligence manageror worse, not trusting him. Basic rule; If you don't
trust the guy running your intelligence operation, why is he still
working for you? Keeping your intelligence people in the dark
makes as much sense as keeping your lawyer and accountant in
the dark.
Never accept an intelligence task where your client or bass
won't or can't let you into tha inner circleunless you really
need the job.
Define the mission and define your intelligence needs according to that mission. What was our client asking us to do? Well, he
knew the product workedit had been tested over and over again
and it did what it was supposed to do. He knew a lot of people could
save a lot of money using the product, but he also knew that hardly
anyone would buy the product.
The mission: Explain why people wouldn't buy the product
and what could be done about it.
The first part was easy. The problem was no one was buying the
product because the company had no hard proof that it worked,
only assertions. The industry was filled with sleazy hustlers who had
developed a reputation for making fabulous claims that never
panned out. The issue, therefore, was how do you prove that this
product works?
Our mission definition looked like this:
MISSION
1. Locate a pattern for validating technologies similar to
magnetic water treatment.
2. Determine what, if any, serious scientific work had
been done in the area of magnetic water treatment
that could be drawn on for validation.

jtting It Together

63

3. Determine if there was actually any interest in a product like this if validation were achieved.
4. Estimate the cost of validation and the probability of
successful validation.
In short, we were going back to the drawing boardto what ought
to have been done in the first placeto see if the project could be
quick-started.
At this point it was a knowledge problem. The investor didn't
know things: Mainly, he didn't know how to get validation. He
didn't know how much, if any, scientific information backed him up
or even if there was any real interest in the product, save his own
intuition and impressions. And he had no idea of the sort of costs
involved. What he needed was information. For example, if he
knew which scientists were involved in the field, he might he able to
cut a deal with them and make them his research and development
team. But first, he needed them identified. Intelligence comes in
early in the process, defining the parameters. It then hands off to
others.
This was, therefore, an intelligence project focused on technology. Our job was to find out about the science behind the
process: What science had been done and who had done it? Also,
we were to determine how scientific research in this field could be
used to reverse the bad image of magnetic treatment of water,
thereby convincing people that buying our client's product was a
good idea. What we were doing wasn't quite science and it wasn't
quite marketing; it was a littie of both.
^

The intelligence professional gives the appearance of being a

dilettante. The appearance is accurate. Easiness intelligence
specializes in the general.
The intelligence methods part of the mission was now defined:
Find out how new products with radical approaches to problems
become validated. With that in hand, we could define our tasks and
get on with the intelligence operation.

64

THE

INTELLIGENCE

EDGE

Step 2: Identify the Knowledge Base

The truth of how much the corporate executive actually knows rests
somewhere between a desperate sense of ignorance and smug selfassurance. It is critical to understand exactly where you stand
what it is that you already know and what it is that you need to
find out.
Businesses are tremendous repositories of information. Every
employee, every mail delivery, every telephone conversation contains potentially valuable information. But most companies make
no effort to retain and organize this information and, as a result,
much of the information that passes through a company is never
captured. This is not dissimilar to accounts receivable casually tossing aside checks at random. Failing to capture information means
that, when the need arises, you will have to go outside the company
to find what a few months ago you had in hand.
We were once involved in a defense-related project where we
discovered that we badly needed information on the precise tonnage of material consumed by a U.S. Army division. The information existed somewhere, but we couldn't find it in the first three
zoneselectronic, paper, or gossip. So we had to go digging
around in Zone 4, the gray zone. We finally found the information
man-weeks latera contact of a friend of a contact knew someone
who had done a study that had never been circulated. There were
other ways to find the information, but this was the one we found.
A few days later, a new accountant on our staff wandered into an
office and started talking to members of the team. It turned out
that the accountant had left the army a month before, having
served in supply. Apart from having a bunch of unclassified army
manuals (always useful) he had the very numbers we had needed
engraved in his soul. We had wasted time and money searching for
something that was in the office down the hall.
This act of stupidity was not repeated. We put into place systems for organizing and accessing our own knowledge base. However, whether or not you have a system to access information, you
must start any project by sitting down and taking inventory' of what
you already know. This is the only way to find out what you need to

jtting It Together

65

! learn. In a one-man show, that consists of quiedy thinking through

and making a record of what you know about a subject. In a larger
organization, the process becomes more complex. But a rigorous
inventory of information is indispensable. Writing it down is important because it enforces a discipline that is badly needed at
this stage.
Take the case of our magnetic treatment of waterthis is what
our information inventory looked like:
INFORMATION INVENTORY

Information from client

1. That the process works
2. Speculation about why the process works
3. A collection of odds and ends on the general theory of
the process
Our own inventory
1. That the process, and the entire field, was regarded as
junk science by most respectable scholars
2. An awareness of the potential importance of the
processif it works
3. An awareness, drawn from our experience in the
defense area, that the U.S. Navy would be a likely customer if the process worked
It doesn't appear to be much to go on, but you play the hand
you're dealt. Note a couple of facts from this list. First, the client
provided the information that the product works. Barring evidence
to the contrary, you've got to take the client at his word. This is not
naivete. He does not need to con you because it is in his interest to
be as accurate as possible. In addition, even if he isn't an expert, he
probably knows more about the field than you do. Trust him on
this. Second, in mastering his field, the client has had a peculiar
and idiosyncratic relationship with it. He was probably not systematic in collecting information, but he surely has more than he is
aware of, so debrief him. Third, our own contribution appeared to

66

THE

INTELLIGENCE

EDGE

be skimpy. We didn't really have any information on the subject, but

we did have an insight derived from our knowledgethat the U.S.
Navy sailed on the ocean and used lots of water in handling dieir
ships. This brilliant insight would actually prove quite valuable.

Splitting It Together

67

p how expensive will this be, what is the potential payoff, and how
probable is it that the payoff will occur?

Step 3: Defining the Knowledge Curve

The obvious is more likely to be true than tho incredible. That's

why they call it obvious, dummy.
Now, here is what we didn't but probably needed to know,
based on the two core questions we had formulated:
INFORMATION REQUIREMENTS

Question: If this product works, why haven't others introduced it before?

Information required

l.Who is or has in the past worked in this field as

researcher or manufacturer?
2. Has anyone used this technology in the past and what
was their experience?
3. Who are the large users of water decakification systems and what are they currently using to do the job?
Question: In this general area, how do new technologies
become accepted?
Information required

1, Identify other processes in the same area as this one,

particularly one that has come in from left field, and
find out how it managed to get accepted.
2. Find out from potential users what they would like to
see happen for them to get comfortable with this technology.
At this point, we pretty much knew what the mission was, and
what questions we needed to have answered. Now we needed to
know whether it was worth our while to answer these questions

Defining the position on the knowledge curve is the crucial step

that makes or breaks an intelligence operation. What the knowledge curve does, as we discussed in the last chapter, is define the
relationship between money, time, and information. How badly do
you need the information? How much can you afford to pay? How
much is it wordi? Some of the answers are a matter of choice. Some
are forced on you as a matter of realityeconomic and intelligence. Defining your position on the knowledge curve is as much
art as science, but it is never casual.
There are three sequential issues to address in positioning
your project.
1. At what point must you act? How much time is there? How
long do you have to gather and analyze the intelligence?
2. How much can you spend? What is the budget? What is the
potential return?
3. In the segment between these two variables, where is the
optimal position to shoot for? In fact, can the job even be
done within the specified amount of time and cost?
In this case, the reaction tinie was set by the financial and legal
position of our client's company. He had about two months in
which to liquidate and litigate or to put in more money. So, we had,
at most, five weeks to come up with the information he needed to
make his decision, giving him a week to absorb the information,
then about two weeks before the drop-dead date to select his
course.
There are two types of curves that have to be distinguished,
one of which might be called the discrete operations curve. As the
name implies, this curve applies to an operation that has a discrete
beginning and end. Here the starting point of the curve is the

68

THE

NTELUIGENCC

EDGE

moment the operation is undertaken, with the closing point being

either the final, usable date for the information or the maximum
cost line. The other type might be called the continual monitoring
curve. Here there is no starting point but rather an ongoing monitoring of the environment. In this case, the curve refers to the
amount of time between the event's occurrence, the acquisition of
information, and the amount of time needed to respond.
In this case we had a discrete operation with a starting date.
The obvious questions were:
How much time was there before the decision point?
How far in advance did the user need to have the information in order to reach a conclusion?

utting It Together

69

tcould be avoided by drawing on the experiences of others

collected by effective intelligence.
At the other end, assume that the entire start-up was a failure,
but diat you would not have invested in it if you had had the information that the entire project was doomed. Assume that 20 percent
of all start-up investment is wasted on errorsan intuitive but conservative investment. Then, the most conservative expenditure on
intelligence would be 10 percent of 20 percentor 2 percent of
total investment. At the far end, any information that avoids a total
write-off is worth $1 less than the investment. Obviously, this is way
too high, since spending intelligence dollars does not create potential value in the way that spending it on investment does. But it does
give us a range for expenditures of investmentssay from 2 percent on up, with the scale rising as the potential risk rises.

How much money was available for the project?

The hidden questions include how much flexibility was there in this
time frame and how much more money was our client willing to
spend in order to get answers earlier?
How much was this product worth? Current antiscaling technology relied primarily on chemicals added to the water, which was
costly and did some damage to the containers themselves. But that
technology worked and people were comfortable with it. So while
the potential payoff for the new technology was large, so was the
risk. This was in every way a high-risk, high-reward situation.
What had our client already spent on professional services,
such as lawyers and accountants, in trying to keep this company
afloat? How much would he have saved in legal fees if he had had
decent intelligence in the first place? How much would he have to
spend on lawyers in liquidating his company?
Remember, the purpose of intelligence is to take advantage of
the knowledge and experience of others, avoiding the extremely
high cost of moving up the learning curve on your own. It is a costavoidance, risk-control investment. Another approach to calculating a maximum expense level is to estimate the amount of money
you could reasonably expect to spend in making mistakes in a startup company, and assume that a certain percentagesayjust 10 per-

Intelligence is cheap thrillsliving vicariously through other

people's foul-ups. Compare the price of admission to the cost
if throwing the party yourself.
Now what is the payoff? This can't be quantified, but it is
potentially very large, as is the probability that no payoff at all will
occur. Now, assume an investment of $3 million, which in various
ways had been estimated as necessary to move this project along.
Assume that there is a 10 percent chance of a $300 million payoff
a hundredfold payoff on a long-shot investment. An investment in
intelligence ranging from 2 to 10 percentthat is, from $60,000 to
$300,000is certainly reasonable, with a 0.5 percent investment, or
$15,000, being the bare bottom.
What was the client going to get for this money? Everything we
could find out legally on the technology of the magnetic treatment
of water, the key players in the field, their experience, and so on.
The client had to clarify a space and time problem. How deeply did
he want us to go, geographically, in terms of zones; and how quickly
did he need die information? The latter question was simple to
answer: He needed it in five weeks.
The geographical question was more complex. Like many
business decisions, this one could be decided on the fly. Having

THE

70

INTELLIGENCE

EDGE

fixed the time, and the initial investment in the intelligence work at
about $30,000, we would be in contact with the client as the operation was carried out, giving him the option of whether to invest
additional money for additional information.
Step 4: Targeting the Zone

There were two clusters of information that we had decided we

needed in Step 2. The first was information on the technology itself.
The second was information on the validation process. These were
two separate but closely linked issues. We really didn't know much
about any research that had been done in the field. Even our client
didn't know of extensive scientific work in the area. This struck us
as odd. Scientists will study anything. They'll study the hell out of a
wart on an antelope's butt. It seemed to us extremely doubtful that
someone, somewhere had not been given a research grant to study
this. One of the great benefits of university tenure is the publish or
perish system. Every professor wants to get tenure because it allows
him to study subjects such as warts on an antelope's buttunder
the cover of academic freedom. In order to get tenure, and later
promotions, raises, grants, and other goodies, professors publish
thingsin fact, they'll publish just about anything, creating an
absolute avalanche of incredibly useless stuff. The authors of this
book won't bore you with examples of their own contributions to
this sludge pile via the noted Journal of Irrelevant Studies.
Howeverand this is criticalwithin this pile are absolute
gems. Buried in this pile is an articlein fact, a whole string of articles arguing with one anotheron virtually any subject you can
imagine. And once you find one article, you'll find that the helpful
professor has done all your work for you. His footnotes and bibliographies contain practically everything ever written on the subjectthat's how he proves he's done his homework. Those articles
also contain everything he missed. Find just one article on your subject and you've found them all. So our first task was to target Zones
1 and 2: electronic databases and the good old library.
There are over 233,400 professors in the United States alone,
studying every subject under the sun, every one of them

utting It Together

71

publishing or perishing. Steal an idea from Microsoft and

prepare to die. Steal an idea from a professor, and it's called
research. But, as with diamonds, you've got to shovel a lot of
gravel to find the good stuff.
Performing the first taskfinding information on the technology itselfproved easier than the second task. Validation was a
more amorphous topic. What did it mean, after all? Well, for us,
what it really meant was how can we convince people that this gadget works and that they should buy it? Somehow we had to figure
out how new technologies got to be taken seriously. The problem
we had was that, unlike most new technologies, this one had
acquired the reputation of being fraudulent. We needed more than
ordinary validationwe had to dispel the reputation of fraud surrounding it.
On reflection, of course, we realized that this reputation was to
our client's advantage. The big guys, like Westinghouse or General
Electric, had not jumped into the game because the conventional
wisdom was that magnetic water treatment didn't work. No one was
motivated to try to overcome the reputation as they weren't hurting
for business. That's what gave our client his chanceif he could get
it validated, if it really did work.
We needed to find out what a big customer would require in
order to buy this product. The biggest customer we could think of
was the U.S. government, and in particular the Department
Defense (DOD). What is particularly nice about the DOD is that
they have rules on procurement; not clear rules, but they are, nevertheless, rules. If we could discuss this with someone, he could
explain to us the rules under which this would be validated. And, of
course, if the DOD bought something, it was validated. On the
other hand, if you could land a Defense contract, who cared about
validation? But nowhere in the tens of thousands of pages of federal
procurement rules could we find rules on this. So we needed to go
deeper, to Zones 3 and 4gossip and the gray areato find out
whether or not the DOD had ever done studies on validation of, or
bought, this technology.
We might find that the DOD had already tested this idea and
that it was on every ship in the fleet. Or that they had tested the idea

72

THE

NTEL

G E N C E

EDGE

and the court-martial documents were available for the poor guy
who lost his ship because of it. Either way, a foray into the first two
zones was going to be the next step, followed by Zones 3 and 4, if
Zones 1 and 2 justified it.
Always do the simpler work first. It frequently saves you
having to do the hard work.
If we found conclusive scientific proof that the process not
only doesn't work, but that it can never work because it violates the
laws of nature and nature's God, well, we probably wouldn't have to
worry about the validation problem.
Step 5: Acquire the Information
Since we are going to devote most of the rest of this book to
detailed advice on acquiring information, we don't need to go into
it too much here, except as it affects our own little operation. First,
of course, we took a careful look at tbe material we had in hand. It
contained a few interesting articlesall of which contained footnotes. Those footnotes, in turn, yielded other articles from which
we figured we might be able to learn some things. It was an important start.
The most important thing in these articles, however, was not
facts or bibliography but language. Every field and subfield has its
own language. The first task in any operation is to master the languageparticularly the buzzwords that are used over and over
again. In this case, terms like rare earth magnets, flux fields, pressure
gradients all occurred and reoccurred. These were absolute gems for
us, but not because they, by themselves, yielded anything intellectuallyat least not at this stage. The value here was that they gave
us keywords to use in database searches. There were three places in
which we began our search, all in the electronic domain:
When in Rome, order in Italian. Every field has its basic
vocabulary that you'll quickly locate because of sheer
repetition. Those five or ten buzz words are your signposts.

(Putting It Together

73

You don't have to understand them at first; just use them as

minini tools.
Dialog: We've mentioned this before as one of the first databases to use. Dialog is a database of other databases. There
are literally hundreds of electronic databases that are searchable through Dialog. But Dialog has some drawbacks. It's
fiendishly expensivesearches run in the hundreds of dollars and upand you really should be a rocket scientist to
use it. Still, it does contain a wealth of scientific and technical
literature, or at least their citations, all searchable by Boolean
operators, which means that in addition to recognizing
whether particular words appear in the text or abstract, you
can use terms like and, or, but not, and so on. That meant that
we could say MAGNETIC AND WATER AND RARE EARTH BUT NOT
ELECTROMAGNETIC, meaning that we wanted articles in which
the words magnetic, water, and rare earth appeared, but not
those that also used the term electromagnetic. Dialog gave us
hundreds of potential articles to pursue.
National Technical Information Service (NTIS): Everything
owned by the federal government is free to the public. A
little-known fact is that nothing that has been produced by
the government carries a copyright on it. It's there for the
taking. Since that was too simple, the government created
NTIS, as a clearinghouse for information. NTIS, like Dialog,
allows you to search using key words but not Boolean operators, which means that you can't exclude things. The search
is free, but ordering articles through NTIS can run you from
$10 and up an article. The article is free; the search is free.
There's a service charge, but it's still not copyrighted. We
didn't find too much here.
Defense Technical Information Center (DTIC): This is the
mother lode, but it's a little tricky gaining access. DTIC contains everything produced by and for the government in the
defense area, which covers a huge amount of governmentfunded research and development. The problem with DTIC

74

THE

INTELLIGENCE

EDGE

is that unless you are a defense contractor or bidding on a

defense contract, you can't get at it...sort of. But a few
calls, which we'll discuss later, will get you in, at least to the
noiiclassified sections. DTIC contains every study done over
the years for the Department of Defense, and in our case,
it told us what work, if any, the government had done in
this field.
Our electronic data dump led us into the paper domain
Zone 2. A lot of the stuff we found on-line, or from the footnotes in
articles we found on-line, were still available only in paper form for
the full article. There is nothing, absolutely nothing, like a first-rate
library, and a business that doesn't have people who know how to
use a research library is a business that is operating blind in one eye.
Our folks shuttled back and forth to the university library, which
had another service worth mentioning: interlibrary loan. If the
library doesn't have it, an on-line catalog called Online Computer
Library Center, Inc. (OCLC), will tell the librarian every library in
the country that does, and the interlibrary loan system will deliver
it to youthough frequently not as fast as you would like. To avoid
any delays, we maintain links with undergraduates and graduates at
other universities. For pitifully small amounts of money, these folks
will dig up the journals, copy the articles, and fax them to you, cutting down turnaround time.
We told you that intelligence work wasn't sexy. But oar society
is blessed with poorly organized, generally inaccessible
archives containing the treasures of our civilization. Willie
Sutton robbed hanks 'cause that's where the money was.
Intelligence types go to the library.
We piled up the journals and copies and read them. We had a
couple of people on the staff who had scientific backgrounds,
although not necessarily in this area. Expertise at this level isn't
really as important as patience, intelligence, common sense, and an
extremely high boredom threshold. What we were looking for at
this stage was simple and twofold: (1) Who was doing basic research
in this area?, and (2) What was the basic research finding?

lilting It Together

75

Most of the articles were hot airliterature reviews of what

I other people had said, one-page articles by people selling the product, indictments against the people who hustled the products. We
$ quickly separated out the noise from the basic laboratory research
on the subject. That cut things down to size. We found that research
had been carried out in only a few countries: die United States, Russia, Israel, Japan. We also found something interesting. Whereas die
literature reviews were highly critical of the process, the basic
research consistently showed that applying magnetism to water did
indeed cut down on scaling, for reasons no one could explain
theoretically.
The lab resultsfrom some very solid universities and by some
solid scientistsdefinitely showed that something was happening,
although diere was massive disagreement over exactly what. We had
now achieved something important. We had enough information to
say diat our client's intuition was not wrong. Something was going
on, aldiough it wasn't clear what it was or why. It also pointed to a
basic problem: In spite of the positive results in the lab, the noise in
the scientific field remained negative. The key problem was the lack
of theoretical consensus. In the sciences, knowing that something
works is not enough. Having a theory about why it works tilts the
balance.
Lists of research projects from DTIC also told us that for about
fifty years the U.S. Navy had been experimenting with magnets to
cut down on scalingalmost since World War II. Every experiment
had failed. In fact, the coast guard had just concluded an experiment using magnets on pipesit had failed. We now had an anomaly: The process tended to work in the laboratory, but it failed in
field tests with the navy. What was going on?
Step 6: Retask the System

On the chart on page 58, retasking the system appears only once.
This is because we were getting tired of drawing loops. Retasking is
a regular process that repeats itself as often as needed. When you
have completed one set of assignments, you stop and evaluate what
you have, creating a new information inventory. At this point, ours
looks like this:

76

THE

INTELLIGENCE

EDGE

INTERIM STATUS AND TASK REPORT

Findings
1. Research had been going on for fifty years.
2. Research centers were in the United States, Russia,
Israel, and Japan.
3. Lab results were uniformly promising.
4. Field tests with the navy and coast guard were generally
a failure.
Question

Why were lab tests successful and field tests a failure?

Task

1. Talk to the navy.

2- Talk to research scientists in the field.
We had completed our work in Zones 1 and 2. Our task now
was to use the information we had developed to penetrate the fog
we had generated for ourselves. In effect, the purely passive work
we had carried out was over. Now our client had to approve a
deeper foray, using what we referred to as semiactive means, which
carried with it some of the advantages of passive intelligence. This
was still relatively cheap, since it relied on phone and fax to ping
the system in order to locate further information, but it was no
longer purely passive. It could be detected.
This posed a dilemma for our investorone that was
inevitable. One of his advantages was the general contempt the
field received from most people familiar with it. By being skeptical
about it, they left it wide open for someone with enough imagination to see the possibilities and incur the risks. As soon as we went
semiactive, we would be alerting people that someone with enough
resources to ask the right questions was out there thinking about
the business. With luck, we would be dismissed as loons. Without
luck, it could start someone else thinking. The more active we
became, the more likely we were to be noticed. The more we were
noticed, the more likely it was that someone would figure out what

Putting It Together

77

we were up to. That risked competition. On the other hand, we

could not begin to dissolve the mystery of the laboratory versus field
tests without more information, and we had collected all the passive
information available.
This is where the client must make the decision. It's not just
his money, it's his risk. The move from passive to any sort of active
intelligence is an extremely serious business decision. It has the
potential to reshape the competitive environment. In many cases,
the client focuses on the increased immediate cost. It's important to
redirect his attention to the real potential costlosing the element
of surprise. In this case, the need for information, and the presumed protection provided by contempt, decided us on a semiactive attack.
Push the passive to the limits. It's not only cheaper, but it
preserves the element of surprise until tha last passible
moment. The extra time bought could be the difference
between an IPO (initial public offering] and being DDA [dead
on arrival).
We proceeded in two directions. The first was trying to find
out what had happened in the navy tests and what was going on in
the coast guard tests. We began by calling the offices that had conducted the tests. A Pentagon directory is handy and can be purchased. We made call after callmosdy to people who had no idea
what we were talking about. But no call was a dead end. We insisted
on a referral to their best guess,as to who would be of use.
Remember, Department of Defense midlevel officials are civil
servants. They are genuinely committed to helping the public. If
you don't believe that, then believe this: They live in dread of a civilian complaint. Most civil servants go for years without getting a call
from outside their department. So getting one is not a minor event.
Getting a call from a civilian looking for information rather than
trying to sell something is extremely rare. Getting a call from a
civilian who is polite, firm, and seems to know about a study that
was done by his departmentin other words, someone
knowledgeableis extraordinarily rare. You will not be ignored.

78

THE

INTELLIGENCE

EDGE

The person at the other end will try to please you, first because it
breaks up a boring day, and second because he wants you to go away
and making you happy is the best way to achieve this.
On the whole, midlevel officials are more likely than not to be
helpful if you let them know that you are not a crank, if you
are polite, and if you behave as if you expect them to help you.
Make it clear that you're not going away.
We finally found the right official. He was in charge of bilge
water for the navy. No joke. We couldn't make this up.
He knew everything there was to know about the navy's
experience with magnets to treat calcification in water containers.
He was the navy's institutional memory on thisand the coast
guard's as well. They had clamped all sorts of magnets on pipes,
but they had no effect. He mentioned several studies, most of which
we had. He offered to send us some others. He was a delight to
speak to and most helpful. We were polite and genuinely interested
in his job.
He also told us that he wished the damned things would
work, as the navy has all sorts of problems with calcification, and
it costs a fortune to prevent it and repair the damage. But he
also said that the navy was not going to fund any more research
and development on the subject, and that the coast guard trials
were the last as far as he was concerned. He said that from here
on out, the navy would rely on private industry's validation of any
new technology. He named several trade and professional associations with national reflations that conduct tests on new concepts.
He told us that if one of those oversaw the tests, and the tests came
back positive, they would give the product another try. But he said
that he couldn't believe the successful studies we cited for him,
since they had tried these magnets over and over and they never
worked.
Second, we went to our literature and located the leading
researchers in the area. Deciding that our client probably would
not fund a trip to Japanand figuring the Japanese wouldn't talk
to us anywayand that we really didn't want a trip to Siberia, where

'Putting ll Together

79

the Russian research team was huddling, we went to some American

researchers. We found a couple of professors toiling away in a fine
university who blew us away with what they knew. Delighted at getting a free lunch at a Holiday Inn, they told us about the work they
had done in the magnetic treatment of waterand the magnetic
treatment of other fluids as well. When we mentioned the navy tests
they laughed. The navy had clamped magnets on the outside of
metal pipes. What the hell did they expect? Metal screws up a magnetic field. They bad never put the magnets on the inside, allowing
water to flow over them. They went on to explain how and why it
works. And on and on. See, no one had ever listened to them
before.
They had been under contract to several small companies to
do work in the area, they told us. They gave us the names of the
companies. They showed us the contracts. They told its everything
we needed to know, and many things that we really didn't need to
know. In the end, we understood why the gadgets had worked in the
labs, why the in-line magnets used by our client did in fact work,
and why most other tests inevitably failed. We also learned about
the kind of water the magnets would work in, and the kind they
wouldn't work in. We learned about other uses, some of them
mightily weird, some promising. And we learned that we were sitting across from the knowledge base of a new industry.
Step 7: Final Analysis and Evaluation

We now had a final report, that could have looked like this:
FINDINGS

1. Laboratory tests consistently provided positive results.

2. These tests were always carried out with in-line magnets.
3. Navy tests were conducted with clamp-ons, which don't
work.
4. Fly-by-night companies sold clamp-ons or magnets
inappropriate to the water conditions.

80

THE

INTELLIGENCE

EDGE

t
(Putting It Together

81

5. The navy would buy magnets if they were validated by

an industry or professional group.
6. These are the names and phone numbers of scientists
who would be happy to work with you.
EVALUATION

1. This is a valid product, but there is substantial risk,

because further research and development (R&D) is
needed to define the exact circumstances under which
the product will work.
2. In addition to R&D costs, the client will need to underwrite testing and evaluation (T&E) for validation.
3. The danger is that with successful R&D and T&E, a
large corporation will come in with its own product
leaving you to defend your patents against their deep
pockets.
RECOMMENDATION

1. The only protection is to create and dominate a market before anyone notices.
2. Be prepared with full and substantial financing before
beginning.
3. Protect yourself against R&D/T&E failure, then blitz
on success.
Now, we could have handed in this one page and let it go at
that. But clients don't like to pay five figures for one page. So we
padded it out with really nice tables and charts, built an appendix,
went down to Office Depot and bought an attractive coverbut
this was ihe basic report. Either be prepared to blitz this with a lot
of financing in hand, or figure that you'll get eaten alive by a big
company who'll wait until you complete the research and development, see how the market develops, and if they like what they see,
beat you to it. It is important to remember that mondis of intelligence work can sometimes boil down to just a single word, "yes."

There is no correlation between the amount of material provided

and its importance or cost.
|
This left our client with a dilemma: Did he really want to take
I the risk? He approached us with a proposition. Would we take a
piece of the business in return for our working with the scientists,
developing the strategic plan, and so on. Well, it was attractive. And
we were attracted. But this was a line we tried not to cross. It was the
line between collection and analysis on the one hand and operations on the other. We were professional pessimists. Even in the
midst of a successful report, we retained our naturally gloomy cast.
If we were to get involved operationally, we would have to become
optimists, cheerleadersdownright positive thinkers. And then
we'd lose our edge. Pushing temptation aside, we declined.
We are, as a company, institutionally pessimistic. Businessmen
are institutionally optimistic. We cross that line at aur peril
and that of our clients. This is why some of the authors of this
boak are intelligence people and some run businesses. It gives
balance.
Step 8: Archive and Database

The project is not completed until every bit of information you've

gathered is archivedput away somewhereand databasedorganized in such a way that it is accessible. This requires discipline. By
the time you're finished with a project like this, you are usually
utterly bored by it. The temptation is to chuck everything into a
closet, lock the closet, and lose the key. It is often assumed that
you'll never need the material again. It is absolutely surprising how
often such material is needed under the most unlikely circumstances. For example, we never expected to be writing a book on
business intelligence and never expected to use this as an example.
The reason we could, and did, was that we shoved everything else
into a closet and can't find the key anywhere. This information we
archived and databased.
Archiving and databasing are made easier by the fact that so
much of our. information originated in electronic format or wound

82

THE

INTELLIGENCE

EDGE

Pputting It Together

83

up in electronic formatafter we reworked it, scanned it, or wrote

it up. We will be dealing extensively with the database problem in
coming chapters. For now, we will assert that our archives and databases are our institutional memory, and that, as the years go on, we
need to do less and less because we have these archives to draw on.
They must be part of any company's intelligence and knowledge
management program.

CONCLUSION

In taking you through the intelligence cycle and the story of the
magnetic water treatment, we wanted to pause and give you an
overview of the intelligence project. This is, of course, a particular
and idiosyncratic project. But that is the nature of every intelligence
project. We found that we could not think up one that was both useful and generic. There is a generic process, but not a generic project. Every business intelligence project is idiosyncratic.
That is why the process itself is so important. The process,
which obviously is not meant to be followed slavishly, is designed to
provide a method that can be used as a general template. And it is
designed to focus attention on some of the issues and trade-offs of
the intelligence process. Now we will examine various types of information and how to find them.
And what happened to the product? Our client understood
clearly what we were saving and bailed out. What we had said was
that there was real potential and lots of riskand that if you were
going to be in for a penny you had to be in for a pound. He decided
that he would rather put in pounds somewhere else, and he did.
We regarded this as successful work. More often than not, successful intelligence provides the decision maker with information
that dissuades him from doing something that he is not really comfortable with. We had shown him the extent of the needed commitment, he evaluated his interests and resources and decided not
to go. Our success lay in showing him what the total cost would look
like before he took the plunge. In many business ventures, the total
cost emerges only over the course of time, frequendy wrecking not

only the project but the finances of the principals. One of the
things intelligence does is save you from mission creep, the nasty
process in which a small hobby turns into a man-eating bear. Intelligence work is frequently more defense than offense. That's good
enough for us. And the check cleared.

[Find Yourself

FIND Y O U R S E L F
Tapping Internal Resources

The information explosion has left companies with a crisis of confidence. Companies of all sizes are aware of the enormous amount
of information available, and are terrified of making the wrong
decision because they missed an important fact. In a sense, this crisis of confidence was inevitable. The growing availability of information caught most businesses unaware. It was as if the sky had
suddenly starting raining information, but there were no buckets to
be found. Obviously, a three-man team with a makeshift bucket
would grab more money than a hundred men without a clue as to
what a bucket was.
The first reaction of many businesses was a stopgap measure
hiring consultants. When the Berlin Wall came tumbling down,
opening up the former Warsaw Pact countries to democracy and
Western investment, major multinationals immediately went seeking information about new and exciting opportunities. Most of
them used major consulting firms in the belief that they had more
intelligence and more insights. In many cases this proved not to be
the case; in fact, the consulting firms were busy hiring former KGB
officers or journalists to provide them with knowledge, much of it
out of date. There was also a plethora of conferences, but few of
them offered anything new. In the end, businesses would have to
develop bigger and better buckets of their own so that the information flowing into a company could be captured and stored. And
84

85

'most important, a system would have to be put in place that would

allow that information to be sorted, accessed, and turned into
knowledge.
Any business intelligence system must begin by utilizing the
information within the company that results from the normal
process of doing business. This is not only the most cost effective
information, but it is also the most precious and noureplicable
such as a salesman's knowledge of another company, or a scientist's
knowledge of a technology. Information is constantly pouring into
your business, but most of it is lost, tossed away, or forgotten. In
odier cases, the sheer pressure of work on senior executives and
entrepreneurs means that they have no time to keep a proper
record of even significant in-house meetings.
The first stop in shopping for information is at home. Most of
what you need to know is there. Finding material you already
awn, however, is sometimes more difficult than buying it again.
Finding information may be as easy as checking a computer
file, a desk drawer, or a Rolodex. But if no one knows it's there, it
may as well be on the other side of the moon. The larger the company the harder it is to locate the person who might have the
information needed and also the knowledge to interpret it. The
foundation of any intelligence system is a structure for making certain that information available inside a business is made accessible
to decision makers.
Several examples come to mind:
*
A company we know in the chemical industry had recendy
divested itself of a division that was a large part of the company. Being flush with cash, it was time to consider new
acquisitions inside the industry. Obviously, the acquisitions it
made would determine the focus of this company for a long
timeit was a defining moment. Although the company was
stocked with thousands of people who had done nothing but
eat, drink, and sleep chemicals, the company turned to outside consultants for advice. None of the information, knowl-

86

THE

INTELLIGENCE

EDGE

edge, or insight of these outside consultants came close to

equaling the huge amount of information and expertise they
had within their own company.
i Another company wanted to present itself to stock market
analysts. It needed to maintain its share price in the light of
some potentially unfavorable short-term results. So it decided
to put together a very nice visual presentation, and hired PR
consultants to do it. However, in getting the material on the
company, the consultancy had to undertake a detailed internal study because senior management did not have enough
information in hand to provide the PR firm.
Yet another example was the subsidiary of a public company
that heard that a much larger organization on anodier continent was seriously interested in a strategic alliance. The chief
executive of the larger company, who was something of a
celebrity, planned to visit with them. Over the years, the
smaller company had amassed a great deal of information
about the potential partner, most of it gained from a longterm, close working relationship. Yet when the day came to
dig out this information, it had all been filed so chaotically
that it could not be found, and executives had to engage in a
mad scramble to learn the essentials.
There are always good and obvious reasons to go to outsiders.
Among the better ones: to fill some defined areas of knowledge lhat
your own company lacks; to serve as a reality check on your own reasoning; to deal with matters that involve extremely sensitive issues
that must be kept secret within the business, such as a decision to
fire everyone and shut down. But many businesses turn to outsiders
for baseline information and knowledge. They are not filling in
blanks or having a "B-Team"a second teamto check their
answers. They are using outsiders to provide information and
knowledge on a regular basis. This is sometimes necessary because
the company has downsized to the point of losing its own knowledge base, and they now have to rent an expert where previously
they had one on staff. The problem of the outside consultant is a

Find Yourself

87

major issue in business today, particularly with downsizing. Internal

intelligence systems may wind up cheaper in the long runand be
more effectivethan hiring outside consultants.

CONTEXT AND NUANCE, DIFFUSION

A N D DISSIPATION

Every member of your staff, including yourself, is an intelligence

gatherer. Indeed, your staff is the most efficient gatherer of intelligence, because it does so simply in the course of doing its jobor
even when not officially on the job. A salesman having drinks in the
hotel bar may not be selling much, but he may be picking up
invaluable information. E-mail, phone calls, lunches, trade journals, high school class reunions, and illicit affairs can all cut into the
amount of time your staff devotes to their primary activity. But diese
activities can provide a critically important stream of intelligence
for your company.
Looking for and finding information is not nearly as easy as it
might appear. There are two problems in retrieving information
from within the company.
Problem 1: Not recognizing the significance of information.
A mail-room clerk at a local bank has a girlfriend who works in
accounts receivable in a restaurant supply house. She calls him and
says she has to work late that nightshe's been ordered to generate a list of all outstanding receivables by the morningso she
won't be able to go to the body-piercing parlor widi him that night.
Bummer. Now, the local bank was working with a client who was
preparing a buyout offer for the restaurant supply house. They had
been working quietly, waiting for the right moment, not in any
rush. The banker spearheading the project would probably like to
know that the company needs to generate a financial statement by
morning, including all receivables. This is not an everyday event.
The stealth buyout is about to get blindsided.
The mail clerk is oblivious to the significance of this information. He has no idea that the bank is participating in a financing

88

THE

INTELLIGENCE

EDGE

package for another group. All he cares about is his girl and body
piercing. The vice president managing the deal doesn't know the
mail clerk, and if he did, it wouldn't occur to him to ask the boy
how his girlfriend is doing. So here we have a critical piece of information in the hands of a company employee who hasn't the slightest idea that it's important, and an executive who badly needs to
know this fact but is not in a position to issue an urgent plea to all
employees for information. First, this is a secret bid and asking
would mean blowing confidentiality, and second, he would be deluged with information and wouldn't be able to extract what was
important from the mess.
This is, in a sense, a classic problem of intelligence. The information is in the hands of a company employee, but light years away
from the decision maker.
Problem 2: The important information can be lost in the diffusion of
gossip.
Here is the paradox. Gossip is an extremely efficient method
for gathering information but an extremely inefficient way to distribute it. Gossip basically represents the diffusion of information in
a nearly undifferentiated soup. The mail clerk's conversation mentioned the late-night mission in accounts receivable, gossip about
some friends breaking up, night classes at the local community college, and a report of an upcoming concert by Night Crawlers. And
this is only one conversation of many in a day. A tremendous
amount of information is transferred with very little overhead.
The inefficiency cuts in during the diffusion process. As you
would expect, what is extracted depends on the interests of the discussant. Of the list above, the only thing the mail clerk cares about
is the inability of his girlfriend to go and pierce body parts that
night. So when the gossip is passed on from the mail room, the
most important news, from the company standpoint, is lost, filtered
out by the mail clerk's indifference. In the end, the low-overhead
collection process becomes a near total loss to the company.
The problem is not in collecting information, but in organizing and controlling it once it comes into the company. Consider
some examples:

nd Yourself

89

Every day, hundreds of ads are dropped on people's desks,

delivered without effort to the business by the U.S. Postal Service. As a rule of thumb, almost none of the mail is of immediate interest to the recipient at the moment he gets it, so he
dumps it in the trash. Yet almost all of it is of interest to
someone in the company. One day, we needed a program
that would allow us to show graphically how a decision tree
branchesdown to a level where there would be hundreds
of thousands of nodes. We searched the Internet, called
friends, pinged all the zones we could think of. It took us
nearly a week to find a software package that would do. It was
FedExed to us. As we were opening it, one of our staff looked
at it and said, "Oh, yeah, that one. I got a flyer on that a couple of weeks ago. Looked interesting." We let him live.
Every day, members of your company make phone calls. Each
phone call provides information soup to that person. The
more frequently a number is called, the thicker the soup, as
relationships develop and information not directly related to
the issue at hand is discussed. As important as information,
context and nuance are created through these conversations.
But content is retransmitted randomly, as interoffice gossip,
stripped of its subtletiesif it is passed on at all.
Departments within companies regularly take on projects
focusing on some aspect of a prpblem. For example, our
company's accounting crew once studied ways to bring down
our extremely high, on-line service costs. They looked at
what each service offered in order to locate the lowest cost
provider or find acceptable substitutes for higher-priced services. A while later we were asked by a company to study their
on-line information providers to make certain they were getting the best value. The research department was tasked with
the study because it knew a great deal about the services.
However, no one in research knew about the extensive
accounting study, even though they had followed their recommendations, with the result that loads of man-hours were
wasted in reinventing the wheel. All of the information was

90

THE

INTELLIGENCE

EDGE

already within the company, along with a cogent analysis.

Gossip hadn't alerted research about accounting's knowledge, and unnecessary laborand lost profitswere the
result. This wasn't supposed to happen to us. We learned
from it.
A company for which one of us once worked needed to invite
fifty top European industrialists to a dinner. We knew their
names and companies from a publication, but not their
addresses. We did not own a directory, though there might
have been one somewhere in the head office. Instead, we
sent a researcher to a business library for half a day to copy
out the addresses, whereas within ten feet of the secretary
preparing the invitations was a computer terminal with an
on-line European corporate database.
Each of these problems derives from an inherent inefficiency
of communications within a company. Efficiency requires a division
of labor in which there is as little overlap as possible. In a well-run
business, people are hired to perform fairly specific tasks, and most
employees are too busy doing their own jobs to do anyone else's.
The result is minimal communication. Most communication is
regarded as leisurenecessary overhead but not essential to the
job. As a result, the efficient and focused organization tends to
duplicate effort, reinventing the wheel regularly because of their
very focus and efficiency.
Indeed, the farther apart two individuals are on the organizational structurevertically or horizontallythe less likely information is to be successfully transmitted. Consider the example of the
mail-room clerk and the vice president. The odds of the two of
them meeting in any given week is minimal. The odds of them
exchanging information is almost nil. The probability of the mailroom clerk having sufficient context to recognize the importance of
his data and taking the initiative of transmitting it is similarly
unlikely. There is even a low probability of two widely separated
executivessay, the director of marketing for one set of products,
and the personnel director at another company siterecognizing
and transmitting information laterally.

I Yourself

91

As a result, information comes into the company and dissifpates. Now, there is no perfect solution to this problem. Ideally,
f there would be a magic box into which everything known to everyone in a company would flowwith no additional cost to the company. Questions could be asked of the magic box and an answer
would be provided. But the perfect magic box is not going to be
created. The probability of the mail-room clerk having an important bit of information in a year is minimal, so die cost of debriefing him would probably outweigh the benefits produced and would
cripple the mail room. Nevertheless, the magic box idea is useful as
a tool for defining the problem and posing some less-than-perfect
solutions. The focus of such a system is not on collection but on
retention and distribution. This is the heart of the business intelligence problem, just as it is the heart of the national intelligence
problem. Critical information is either not recognized as important
or not transmitted to people who could act on it.
Consider one of the most famous examples. There were officials in Washington who had clear indication that the Japanese were
going to attack the United States. These were not the highestranking officials, but relatively minor intelligence officers. First,
they had trouble convincing themselves that the information was
valid. Then they had the devil of a time getting the attention of
senior officials in Washington. Then these senior officials were
unable to transmit sufficiently clear orders to Pearl Harbor. The
internal communication system, from receptors to decision makers
to key outposts, was simply not geared for the rapid internal transmission of information.
Focusing on the collection system, spies and satellites and the
other paraphernalia, is worthless unless the information they
gather can be sent to the appropriate points, analyzed, and distributed. Frankly, if we had to choose between tripling the efficiency of
collection or doubling the efficiency of internal dissemination,
we'd easily go with the latter.
In creating efficient intelligence, increases in internal
retention and distribution increase efficiency geometrically:
Increases in collection increase efficiency arithmetically.

THE

INTELLIGENCE

EDGE

Imagine a particular bit of information and ask this question:

Is diat information more valuable in one person's possession or in
the hands of five people? The answer is obvious. Now imagine five
bits of information in the hands of one person or one piece of
information in the hands of five people. Wbich is worth more? This
makes the situation more complex. A single bit of information may
have value, but its value increases as its context increases. Take the
case of our mail-room clerk. The informationabout his girlfriend
working latehad very little value by itself. It became valuable only
in die context of other informationonly in the hands of someone
who had perspective on the matter at hand and the authority and
interest to do something about it. It is extremely difficult to predict
the precise matrix in which a particular bit of information will have
value. There really was no way of knowing, just by looking at that
information, where value would be found.
This means dial the probability of any bit of information being
fully exploited increases as it is embedded in the largest number of
contexts. One bit of information in the hands of five people yields
five information matrices to provide information. Five pieces of different information in die hands of one person provides only one
matrix to provide contentalbeit a context enriched by five pieces
of information. Let's say that five people have equal amounts of different information, with a value of N. Let's say that a piece of information is found with a value smaller than N: X. Now, in a very
simple model, let's say that the potential value of X is its relationship with each individual component of N, which is to say that its
value is in linking it to every other bit of information. This is not
altogether true, because not every link will prove fruitful; but on the
other hand, this ignores the increased values of links with combinations of information.
That would mean that die value of any bit of information
inside a given matrix would be XNan exponential value. Exposing
to multiple matrices, increasing the number of contexts it appears
insay, giving the information to five people, all witii different
informationwould create a value of XSK. Giving one person five
pieces of information, on the otiier hand, would create a value of
5XK. If graphed it would appear as follows:

93

[Find Yourself

BIST1IIUTEI
IHFBRMItTIIN

ACCUMULATED
IHFIIMITIOH

MLIECTES IHF1RMITIOH

AMOUNT OF INFORMATION COLLECTED

All of which is an extremely pretentious way of saying that letting a lot of people look at a litde information is better than having
one person look at a lot of information. The key is in the efficient
distribution of informationeven more than it is on die collection
of information.
Companies naturally accumulate information. The value of
this information is rarely realized because the information realizes
its full value only when it is embedded in other information. The
more information matrices in which it is embedded, the more valuable the information becomes. Most information collected as a byproduct fails to accumulate anywhere in the company. It is collected
by a person who does not know enough to understand its value, and
its life ends there. It rarely diffuses to enough people to maximize
its value.
As a result of this, most companies look for information outside the company, not only for necessary information that is
unavailable in die company but for information in general. Most
companies might accept the premise that the people with the most
substantial knowledge of the business and its needs already work for
the company, but they also believe tiiat internally accumulated
information is so diffused that it cannot be harnessed efficiently.

94

THE

INTELLIGENCE

EDGE

Therefore, rather than building internal systems for distributing

accumulated information, they turn outside.
This costs more than simply the cost of repurchasing information. It carries an additional price. Outsiders may have a richer collection of exogenous factsabout the industry, business in general,
politics, regulations, or whateverthan insiders do. But outsiders,
by definition, lack the same information set as insiders; their matrix
is never as rich as those possessed by insiders. Why, you might ask,
is this important? The crucial factor here is that context and critical
nuance are missed. The reason this appears necessary to managers
is the realization that this advantage derives from their focus on collecting information on these external events. Managers are therefore willing to buy this informationand analysisoutside,
thereby sacrificing the finely textured context needed to apply it to
the particular company. As a result, there is rarely synergy between
the outside consultant's knowledge and the precise reality of the
company. At best, two information setsone from the consultants,
the other from the executivecome together on one or a few
desks. Most consultant reports land in a file cabinet, and executives
continue to fly blind.
The key problem has been the inability of businesses to recognize the amount of information being collected and dissipated,
along with a sense that it was impossible to accumulate and distribute this vast amount of informationor that the costs of accumulation and distribution, both in terms of time and money, would
outweigh the benefits. Information controls, we would argue, are as
valuable as cost controls or any other efficiency matters. The key is
not to overreach. The goal is not complete, but increased, accumulation and distribution. And, as in many things, the first and simplest measures can show the greatest returns.

COLLECTION,

INFORMATION,

KNOWLEDGE

We've all seen movies like 2001 and Three Days of the Condor, or TV
shows like "Star Trek." Each featured a magic box. You could ask virtually any question. It knew everything, including what it is that you
meant when you asked a question. For example, on "Star Trek," a

I1 find Yourself

95

I question such as "What is out there?" would be understood by the

| computer as referring to that strange object coming at you at the
I' speed of light, and not the litde particles of dust surrounding your
I ship or the stars all around you or the almost limitless other objects
that "out there" could refer to. In reality, the magic box had the
understanding and ability to provide context and nuance that a

human being has, with the addition of a perfect memory and

unlimited knowledge.
This defines the problem perfectiy: Human understanding is
indispensable for turning information into knowledge; on the
other hand, human beings suffer from insufficient information and
memory. A single human being cannot possibly absorb or store all
the information needed to answer the questions put to the magic
;. box. To put it in computer terms, a human being may have enough
RAM, and even powerful enough programs to manage information,
but he doesn't have nearly enough disk space, nor enough input
devices to store the information.
Every company needs a magic box, but it is impossible to create a machine that knows everything. It is even harder than impossible to create a machine that can understand the question
perfecdy and can respond unerringly. But it is useful to keep the
magic box in mind as a touchstonean unreachable goal that can
nonetheless serve as a standard and a measure.
If you wish to wait until artificial intelligence comes along to
solve this problem, we suggest that you find i bobfay to while
away the time. We suggest watching glaciers move.

That does not mean we have to tolerate the perpetual dissipation of information. Certain computer-based and human-based systems can be instituted to decrease, if not eliminate, the dissipation.
In our example of the mail-room clerk and his girlfriend, it is highly
unlikely that any system can be devised that would make bringing that information to the vice president's attention anything
more than a matter of luck. But creating routines for managing
accumulated data can decrease the amount of luck needed and
increase the probability that the right information can reach the
right person.

96

THE I N T E L L I G E N C E : EDGE

For many, indeed most, of these tasks, we will want to use computers. But before we leap to computers, it is important to think
about the process without computers.
One solution to this problem is to encourage everyone on
your staff to think about the value of things that they've learned and
voluntarily pass the information to those they think might be interested. This is certainly something that should be encouraged. Nor
is it a trivial improvement. Creating an atmosphere that encourages
the identification and dissemination of useful information is an
obvious, simple, and rewarding first step. And if you aren't prepared to do that, you might as well forget the rest.
But there is a problem here. It is difficult to recognize important information from the limited standpoint of any single job, particularly those that are uniquely situated to collect information. You
can encourage it, but given the limited amount of context, you risk
a garbage dump. With the best intentions in the world, people lacking context, who take it upon themselves to bombard others with
information they think is valuable, will quickly overwhelm the system with the trivial and the obvious with litde benefit. Sometimes
when we hire new people in our shop, they get enthusiasdc and
start dragging in articles from die front page of the New York Times
or Newsweek. They just don't have the context. They're cutefor
about a week.
This process needs to be brought under control, and this
begins with identifying people in the company who are likely to be
collecting information on a routine basis and encouraging them
to think more clearly about what they are finding. This also requires context, but to a more limited segment of the workforce, for
whom the time and money spent on creating a powerful contextual
sense about the company is likely to pay off. This is due to their
position in the company and the likelihood that they would remain with the company for an extended period of time, as well as
peculiar personal skills that managers can recognizelike being
gossips.
Sane people, by natire or function, are natural collectors.
Identify them and use them.

Yourself

97

Every company has a different cadre that you might want to

identify as prime collectorsit really depends on the type of business you're in. These people share certain characteristics:
Routine heavy exposureextremely high exposure to information and events outside the business in the course of their
work. This refers to people who deal with large amounts of
externally generated material. Salesmen and purchasing people are obvious choices. But a longtime accounts receivable
clerk can have a sense of the ebb and flow of other businesses that is more powerful than even the CFO's sense. People who deal with maintenance and real estate issues are
similar, and the person who goes to the bank every day picks
up the damnedest information.
Multidepartmental activitiespeople whose tasks involve
more than one part of the business. Obviously, in small companies this will include everyone, which is one of the advantages small businesses have over large ones. In large
companies, people who process information from various
departments are particularly valuable collectors. For example, the clerks who process travel and entertainment vouchers are in a strategic position to recognize potentially
valuable information because their position provides context
by definition.
Longtime personnel. Every company, at every level, has
employees who have been with the company for years before
the deluge of information began. These people frequently
constitute the company's internal memory. In healthy companies, these people are recognized and rewarded, quite
apart from the specific job they are doing, for that institutional role. It is fairly easy to spin that role into the intelligence collector role. If anyone can recognize context, these
people can.
So the ideal candidate here is the thirty-year veteran of the
purchasing department who takes care of company-wide needs at

98

THE

INTELLIGENCE

EDGE

all levels, but especially in strategically significant purchases, who

has seen everything, knows everyone inside the company, and outside as well. Getting this guy to pass along everything he learns that
he thinks is of value will obviously be rewarding.
The second best solution is to identify strategic collection
points in the companythose departments or segments of departments through which large amounts of information routinely pass.
In other words, select departments first, then identify individuals
within the departments you would trust with the task of keeping an
eye out for significant information and passing it on. Ultimately,
you do not want everyone in your company acting as a collector.
What you do want is people at strategic points being tasked to act as
collectors.
The percentage of people tasked as collectors within any
department can and will vary. For example, the sales force is inherently intelligence-oriented. They are in constant contact with the
market and are the most sensitive to what competitors might be
doing. You might want to designate 100 percent of this group for
intelligence gathering. In another department, say among routine
assembly workers on an assembly line, the number might be
extremely low or even nonexistent. It obviously varies by the type of
business activities in which they are engaged, The focus must be on
selecting your collectors to maximize the probability of capturing
valuable information while minimizing the burden that collection
places on your main business.
These people will constitute a relatively small fraction of the
workforce who will be doing a value-added service as intelligence
collectors. An additional selection criteria will be trustworthiness. They will have to be briefed regularly about what is happening inside the company, and trained and tested to make
certain they keep their mouths shut. They will need this briefing
in order to recognize what is and what is not important. They can
also be tasked to carry out additional searches in their area of
expertise.
You cannot hive an intelligence service if you don't trust
anyone. Certainly, some people will blab about what they've

I Yourself

99

leaned. Bit the amiunt of information you collect will

atweigh thit blabbing by several orders of Magnitude.
Information gathering for intelligence is a cultural shift,
requiring a redefinition of value to include knowledge production
and making everyone part of that process. The cultural shift is
needed in order to allow employees to see their designation as a
collector as a rewarda new opportunity to add value to their positionsrather than as a booby prize to be avoided at all costs. The
job of information collector cannot be carried out without
rewardswhether these are immediate financial rewards, increasing the probability of promotion, or something quite different.
Collecting information for the intelligence department can be
incorporated into the general demands of the business, just as
many companies assimilated the idea that safety is everyone's job, or
that customer satisfaction is everyone's responsibility. The rewards
can be shared, too. In one bank we know of, workers in every
department are rewarded if someone brings in a new customer.
If you do not use everyone as an information collector, but
choose to designate only certain people, tensions can arise between
those who are doing their usual jobs and those who are doing both
their job and an added intelligence function. This is more likely to
happen in a department where, say, six people are doing their regular job and the seventh is designated collector of information on
top of his existing position. How do you reward the one who is
doing what is seen as the extra work? And how do you keep jealousies from arising in those who feel he is getting die more important, or sexier, role? It is important to recognize that you are not
using this person in a primarily intelligence-gathering position. His
primary function remains what it always was; all he's doing is thinking about what he's seeing in a broader context than his colleagues.
It is important not to overdo the sexiness of the task. He is just a collector of information.
There is a more delicate issue herea paradox that in some
ways cuts to the heart of a business's identity. The more a particular employee derives his value to the business from his role as an
information collector, the greater his value to the intelligence sys-

too

THE

INTELLIGENCE

EDGE

tern, and the greater his reluctance to transfer valuable information

into the company. His value to the company is his knowledge of
what is going on in the marketplace. If a salesman is good, he knows
his customers, their needs, his competitors and their product lines
better than anyone in the business. He is naturally inclined to hold
this information close to his vest, out of fear that by transferring it
into the company's knowledge base, he would undermine his own
value. If the company knew everything he knew, then maybe the
company would decide they don't need him.
This is indeed a serious problem. Businesses that would use
this transfer against their employees are clearly not going to
develop effective intelligence systems. But frankly, companies that
treat employees this way will have much bigger problems than lack
of intelligence. A culture that would treat employees that way
breeds systematic disloyalty throughout the entire structure. Everyone is guarding his own expertise against poaching. The only companies that flourish under this sort of regime are those where
employees see themselves as independent businessmen operating
under the general umbrella of a company. Examples are sales organizations like real estate companies, stockbrokers, or insurance
companies, where salesmen are treated less as members of a team
than as competitors and independent entrepreneurs trading part of
their commission for a base of operations.
Reassurances and incentives must be made to information-rich
personnel in order to induce them to pass information through to
the general knowledge base. One of the assurances, particularly
with people like salesmen, who are performance-oriented and measure performance in precise quantities, must be that the collection
task will not cut into their primary function or that, if it does, suitable compensation will be made. However a company structures
this compensationcontests, bonuses, estimates of percentage of
time spent on intelligence and appropriate paythis drives home
a critical point. In order for an intelligence system to work, its cost
must be rigorously controlled. The system must not become a burden, but be a value generator. It must therefore be kept simple.
Encouraging employees to turn over information requires that
businesses resist the temptation to screw their employees.

Find Yourself

101

Intelligence is a ling-term proposition, and the value of

collectors must transcend immediate advantage.

MINING THE DATA LODE

Imagine that a magic box were a bulletin board by the coffee

machinenot necessarily a bad place to post important intelligencethen the collector would have to write it down, carry it
down the hall, and post it. Each step takes time. The same with loading the information into a database. Now, these steps might seem
trivial, but multiply it by a small fraction of the information rushing
at a business and it rapidly becomes a burden. Let's remember an
inevitable principle of our era: Everything inflates. There is a natural tendency to think that everything is important, a fear that you
might miss something critical. Intelligence inflation is real, and you
want to resist it. That's why the first information you should collect
and sort is the information you've already got.
Your business has perfectly ordinary, prosaic departments: customer service, ordering, maintenance, and so on. Nothing unique,
nothing sexy. Yet, as we've argued, each of these departments, simply in the course of their everyday work, collects information. Your
first task in building an intelligence system is using the information
gathered in the course of this business to provide you with knowledge about what is going on around you. You are not going to be
able to build an actual intelligence system from scratch. What you
have to be able to do is to use existing systems in another way. You
may use computers in this endeavor, but the real key is redefining the mission and indoctrinating the staff with the new ethic of
intelligence.
British Airways is perhaps one of the best illustrations of the
successful collection, retention, and distribution of information
within a company. BA's success has been the creation of a system
that involves all employees and staff in the collection and dissemination of information about their product: customer care in the airlines. Sir Colin Marshall, chairman of British Airways, is a great
proponent of the argument that global communication and technology should not replace the human side in the ability of a busi-

102

THE

INTELLIGENCE

EDGE

ness to practice professional customer care. He states; "The secret

of success will always depend on the creativity that is put into the
use of information."
British Airways now has one of ihe biggest information and
communications systems in the world, with more than 200,000 terminals linked to mainframe computers. The airline employs 2,200
information technology staff, spending on it more than 150 million a yearabout $245 millionthe equivalent of two Boeing 747
aircraft. BA has many different databases, but they are turning
toward data warehousing to create more specific use information
centers to aid in analysis. BA is trying to create a system that is accessible and useful to the customer as well as the company internally.
Prior to its privatization and when in British government ownership, BA's knowledge of its customers and their wants and habits
was almost negligible. After privatization, British Airways introduced a program called Putting People First, where they refocused
the whole company around the customer and his needs. The BA
staff are trained to work with their management to increase sales
and customer loyalty. They collect information on the likes and dislikes of passengers, on their travel habits, and on how they rate their
service on the actual flights. Senior managers get involved in customer-listening forums and listening in on phone calls from customers. Senior managers also discuss customer service problems
with the customer relations department, which has set up a Care
Line for customers to express their satisfaction or dissatisfaction
with their service. It's always easy to listen to positive feedback from
customers, such as when they call to report good service on a flight;
the hard part is listening to them when they're unhappy or irate.
BA takes the altitude that the department of customer relations is the "customer champion" department, and its goal is to
keep customers happy, thereby keeping their business. (k>od customers of the airline receive a birthday card on the appropriate
date sending them the airline's best wishes, and informing them
that, as a reward for being such good BA customers, they are being
given an extra five hundred air miles. Essentially, the marketing
tools have two purposes: One is to keep the customer happy; the
other is to open a line of communication to the customer. This

ind Yourself

103

means that the customer service rep is a prime information collec[ torand one with potentially enough context and nuance to sort
and distribute the information he collects.
Complaints are critically important intelligence. Just collecting
complaints, however, is obviously not enough. It is most important
not to let these complaints diffuse and get lost once they enter the
company. BA has a system to collect, track, and enter information
in its databasepassenger complaints as well as habits and wishes.
Instead of seeing complaints as a pain to deal with, or an area of failure, BA sees the customers as doing it a favor by reporting any problems they have had. How else are they going to know if they have
unhappy customers? They also try to deal with them efficiently,
responding as quickly as possible.
British Airways developed their own analytic side of the company that objectively evaluates the operations part. This B Team,
called a marketplace performance unit, is totally separate from the
marketing, selling, and operating part of the company. This unit
views BAfrom the perspective of the customer, independently judging and measuring its performance. It was recognized that the operational side of BA-the crew and staff who interact with the
customersneeded feedback on their service. But in the service
business it is hard to get reliable data. Apart from evaluation and
feedback from its B Team, BA gets customer feedback by handing
out survey cards to all their passengers and taking random samples
from diose who have arrived at their destination. This gives a customer an opportunity to comment on his experiences flying with
British Airways. The service director is the one in charge of the
cabin crew, and he and the c.rew use the information from the
filled-out cards to evaluate their strengths and weaknesses in meeting their goals for customer satisfaction.
British Airways uses its various clubs as collectors of information. What the British Airways Executive Club actually amounts to is
a very extensive and useful database. The airline knows where and
when you travel, whether you travel alone or with company, where
you live and work, where you stay when you're away on business or
pleasure, your nationality and passport number, the name and
address of your travel agent, the identity of your secretary, the type

104

THE

INTELLIGENCE

EDGE

of car you like to rent, and an assortment of other pieces of information, some of it useless but much of it relevant to route planning
and customer service. The basic data is, of course, gathered from
Executive Club application forms, which can be updated every time
you make a reservation. From time to time, mailings or conversations with a trained reservations agent will elicit more detailed personal information, such as where you would like to go for a
vacation.
A senior media person in a British company received a friendly
letter from a computer posing as a senior executive of the airline,
noting that he had often traveled in one direction on British Airways on any given trip but seldom in both directions. Was there
something wrong? To give him incentive to use BA for both legs of
any journey, the Executive Club offered him double air miles on his
next round-trip. A few weeks later his secretary received an invitation to join the BA executive travelers' Personal Assistant association. This provided a monthly newsletter detailing changes in
scheduling, a special reservations number, and the opportunity to
win free trips.
The airline's success in establishing and using its database system is all the more remarkable because it has been achieved at
modest cost. BA's chairman was astute when he recognized that the
development of new technologies would give the airlines more
competitive ability in real terms, at less cost, than they ever imagined. Everything is centered around the collection, storage, and
analysis of information to provide better-quality service for their
passengers, thereby ultimately increasing the company's profits.
As the British Airways example shows, data already being
created are often primarily numbers being crunchedand that's
extremely useful. What we are suggesting, however, goes a bit
beyond that. What is being tracked here is knowledge before it gets
loaded into computers, by tracking the traces of such knowledge.
This material can serve as the core of a magic box into which a wide
variety of material can be loaded.
Some of it is immediately usable, such as sales reports to track
the sales of your product in various regions and in various product
lines. But there is other information inside your company that you

ipnd Yourself

105

get regularly, use for certain purposes, and let go. For example,
think about your phone bill. You receive regular itemized records
of all your calls, either as a printout or, at your request, as computerized data. Accounting goes over the bills in order to make sure
that it's coming in within budget, that people aren't calling Tibet
on personal business, and to make certain that your phone company is giving you a good deal. This kind of prepackaged information comes in the door regularly and is available in formats that can
be fed directly into your personal computer (PC).
This is information that is generally computerized to begin
with, so there is little or no overhead cost involved in getting it
loaded up. In most companies, these datasets are used only for their
original purposeaccounting, sales, and such. They are rarely
made accessible to people outside those departments. This means
that while the direct utilization of information does occurand
occurs with a high degree of efficiencythe indirect use of information is neglected. Put a bit differendy: We create information for
a reason. But information can provide knowledge over and above
intentions, if it is made accessible.
Consider a fairly simple and straightforward case. One of the
biggest problems in any company is finding someone who knows
someone. Let's say you have decided to try to sell something to a
new company. Obviously, a cold call is not nearly as good as starting
with someone with whom you already have some sort of relationship. Finding such a contact is a critical part of the sales process.
Assume that no one on the marketing team has ever had any dealings with anyone at that company. In a large economy like the
American or European, this isn't unusual. The next step is to find
someone within your own company who has a contact. Sometimes
this works. Sometimes, particularly in larger companies, asking
around doesn't help.
The problem is the same as in our mail-room clerk/banker
example. Information may be present in a company, written in
someone's Rolodex, embedded in someone's brain. The owner of
the information doesn't know of the need for the information.
There may be somebody over in engineering whose closest friend is
in charge of buying the stuff that the marketing people want to sell.

106

THE

INTELLIGENCE

EDGE

The engineer would probably love to pick up brownie pointsand

his friend would at least be prepared to give A hearing. It's a perfect
situation except for the fact that the marketing folks never heard of
the guy in engineering and that guy has no idea that his knowledge
is needed.
Now, one way to solve this problem is to have everyone enter
into a database everybody they know. Sounds good, but it will never
work. Apart from privacy issues, people are just not going to keep
these databases updated. Like many such postings, it sounds good,
but the database would very quickly become stale. Positive feedback
systems, systems that require active, voluntary cooperation by large
numbers of people, simply don't work in cases where there is no
clear use of any given action.
Databases abound. People who bother ta use them are rare.
This is a place where the indirect use of information is important. Question: How is it possible to identify people with contacts in
other companies, without having them actively identify themselves
or without using informal means, like gossip. Answer: Look at the
means that people use to stay in contact with other people. No
rocket science needed: People stay in contact using the telephone.
Friends stay in touch. Modern telephone systems can track even
local calls, either internally or from telephone bills. Check them.
The bills were not created in order to track relationships, but this is
an extremely effective unintended use. If you want to know if someone has any contacts over at XYZ Corp. and asking your four closest colleagues doesn't yield anything useful, then find out the
phone numbers used at XYZ Corp. and check your billing records.
If you find an employee who regularly calls someone over at XYZ,
you've got a pretty good indication that he knows someone over
there. Maybe your employee can help you get in over the wall.
Sometimes it's just a case of pawing through old phone bills. But
since most of the information produced by the phone company and
your own phone system is in digital form already, it can readily be
loaded into a database and mined.
What we are really talking about here is constructing intelli-

Rnd Yourself

107

gence systems out of existing material to track expertise within your

own companyand knowing people in the industry is very much a
variety of expertise. Nothing is wasted more regularlyor more
expensivelythan expertise. People within your company have a
tremendous amount of knowledge that goes far beyond their job
description. Usually, this unexpected gold is found accidentally, or
when you have a particularly aggressive and insightful employee.
Usually it is never found.
The first and most important mission of an intelligence system
is the location of information and knowledge within your own company. That is not as difficult as it may sound. People with knowledge
and skills leave tracks. In this case, phone records can reveal that
someone in your company has a contact in a company you want to
sell to. Those same phone records will tell you if there is anyone
who knows anything about a particular government agency, the
stock marketanything that requires telephone contact. The
phone call is the activity. The phone bill is the direct result,
designed to get accounts payable to generate payment. But that
same bill is a place where experts leave evidence of expertiseand
they can be tracked, with or without computers. The phone bill
allows you to infer things that are valuable.
Phone bills are just one example of traces inevitably left by
experts that can allow you to locate internal expertise. Here are
some other sources of information, found in most companies, that
can be used to track expertise:
Purchase orders, listing who purchased what and where can
tell you who in your company knows about particular technologies. If someone has been ordering a lot of database
software, he might be the one to discuss your own database needs.
Travel and entertainment vouchers, listing where employees
traveled and who diey had meals with. You can use this as you
use your telephone bills, but since the purposes of travel and
entertainment are usually included, this is an even richer
lode of information. By hooking this into a CD-ROMbased
yellow pages, each person's company can be identified with

108

THE

INTELLIGENCE

EDGE

an industry. Run a computer check for everyone who has had

lunch with real estate specialists in Seatde and you might find
someone in your company able to discuss expansions in that
market.
Interview records: Your human resources office is constantly
conducting interviews and constantly collecting information
on people. Entered into a database, the applications jobseekers fill out can contain valuable information. Let's say
you have a sudden influx of people applying for jobs who
previously worked for one of your customers. Worrisome
and definitely something you want to know about. Now, even
four or five applicants from that company can be significant,
but not if they're lost in a fairly large human resources
department. Collecting these and analyzing patterns is not
why you are collecting these files, but they are certainly an
outstanding use.
All of these represent another use for an existing data source.
It's cheap because you're collecting it anyway. The only thing you
are paying for is the interrogation systemthe mechanism that
allows you to ask questions. The list of sources isn't exhaustive. Just
look over the files you maintain in your own business to get a sense
of what treasures you have. Then think through what other information could be inferred from those records and create a system
from a shoe box to a mainframeto track what is going on in your
own company.
So, for example, if we find that John Smith over in information
systems called XYZ Corp. on average four times a month for the
past year, traveled there once, took two XYZ Corp. employees to
lunch, and purchased several of its products, the chances are that
Mr. Smith knows quite a bit about XYZ. Before you make your cold
call on XYZ, sales might want to chat with Mr. Smith. Mr. Smith
might even be tasked to gather more specific information. But
without what we might call an inferential data system (and doesn't
that sound grand for some files in a shoe box), how will sales find
out about Mr. Smith?
The most valuable data in your company, therefore, is the data

' Find Yourself

109

that you are collecting anywayso that it doesn't cost you anythingand from which you can infer other useful facts. The most
important of these facts is always expertise. Tapping into the knowledge base of your company should always be the first step in an
intelligence operation. First, the answer is probably right tfiere. Second, if anybody can find the information you need, it's the one who
already knows an awful lot about the system.

BUILDING THE M A G I C BOX

There are, therefore, two types of important information captured

by businesses. One is direct informationthe kind of information
that can be captured directly by collectors. The other is indirect
informationfrom which unanticipated knowledge can be inferred. Both are being collected. The problem is that information
is not getting to the right people. The collectors frequently can't
recognize the value of the information that they've gotten hold of
and the decision makers don't know what information there is available or who has it in hand. This is not a problem of collection but
of organizing and providing access to collected information. Businesses simply need to continue what they are doing in order to collect it. What they need to change is what they do with the
information after they've collected it.
A systematic means of disseminating information is critical.
The system can be extremely simple or sophisticated, but it must be
able to achieve certain definite.functions:
The information must be disseminated quickly enough to
permit action. It must be well positioned on the knowledge curve.
The information must have some degree of targeting without
losing the exponential advantage we have discussed. This is
difficult to achieve, but important. You do not want to inflict
a daily information dump on everyone, because important
information will tend to get lost or the entire process will be

110

THE

INTELLIGENCE

EDGE

ignored. On the other hand, you want to achieve the widest,

appropriate, distribution of information possible.
The information must be preserved in some archival form, so
that it is retained for consultation over time.
The archive must be queriableit must be organized in
such a way that the information base of the business grows
over time.
A basic, notional outline of the internal business intelligence
would look like the chart below.

QUERYINfi

AND
TASKING

Find Yourself

in

INTERNAL INTELLIGENCE M A N A G E M E N T

There are four general approaches to information/knowledge distribution:

Send everything to everyone.
Have the collector route the material.
Have a central control on distribution.
Create a pull system where information consumers decide
what they want to see.
Each of these systems has drawbacks. The first gives us information overload, in the second the collector makes decisions he
may not be qualified to make, the third increases overhead by creating a central office, the fourth assumes that the consumer will
know what is available before he sees the material. Put simply, what
is to be done with all the information accumulated?
The first, simpler, part is the question of what to do with the
information by those who have sufficient context and sophistication
to be able to recognize its true significance. This is a problem of
knowledge management, since the information has already been
turned into knowledge in the collector's mind. The second, more
complex, issue is what is to be done with the pure information
information that has been collected but has not been recognized as
valuable by the collector.
The first case is easier to manage, since the only question here
is whether the knowledge should be directly forwarded to an appropriate recipient or whether it should be transmitted to a central
point for prioritizing and routing. Although different, these two
strategies are compatible. One of the advantages of a small business
is that the collector who recognizes the value of the information he
has found will be in a position to identify the optimal user of the
information.
In the larger company, this becomes a much more serious
problem. There is a point in the growth of a company where everyone is no longer familiar with everyone else. This normally begins

112

THE

INTELLIGENCE

EDGE

with newer, junior employees being unknown to older hands. At the

next stage, middle managers cease to know other middle managers,
but can still identify senior managementalthough they might
have declining access to them. In some companies, senior management ceases to have direct contact or even full knowledge of other
senior management and their responsibilities.
The larger the company, the harder it is for information to flow.
You can't swap stories with someone you've never heard of.
This process of growing managerial anonymity poses a tremendous challenge to the diffusion of knowledge and an even greater
challenge to the diffusion of information and its transformation
into knowledge. In large companies, the context available to collectors declines, and therefore their ability to transform information into knowledge declines. Most of the input into die system is
raw, uncontexted information. But even where the information is
turned into full knowledge, it is difficult to know who to pass it on
to. And when the collector knows who to pass it on to, it may be
extremely difficult to gain access to that person.
Current managerial doctrine calls for maximum openness in
order to avoid this problem. Sounds good. But we've been there. In
very short order, managers will be swamped with a huge quantity of
information from those who think they understand context. The
informal paths and relative intimacy of smaller companies increases
their understanding of context sufficiently to allow collectors to
screen out the chaffwith informal penalties for eager-beavers who
keep dumping garbage and pretending it's the crown jewels. But in
larger companies, both knowledge and information have to be
managed.
In businesses that are large enough to achieve a certain degree
of internal anonymity, the informal evaluation and transmission of
information and knowledge used in smaller businesses would
quickly create gridlock. Managers are users of information and creators of knowledge. They cannot simultaneously be evaluators, prioritizers, and information managers. Some minimal overhead
system must relieve them of the information/knowledge manage-

Find Yourself

113

ment burden, while allowing them access to needed information.

Indeed, even small businesses need a system for storing and retrieving information, since even the best entrepreneur forgets.
There has to be some sort of organizational system that
achieves three goals:
1. Provides a means for collectors to input information and
knowledge.
2. Provides a system for prioritizing and routing information.
3. Provides an information storage location that can be efficiently queried.
Obviously, this needs to be built around a computing and databasing system that we will discuss more extensively later on. For our
immediate purposes, the issue that needs to be discussed is the
loading processwhat is to be done with information collected by
individuals. Storage, sorting, routing, and querying are all part of
the general system. But what is a salesman to do with a nugget of
knowledge once he's found it?
Mechanically, this has a great deal to do with the type of management system in place. We are talking about a computer-based
system and some sort of information/knowledge database into
which random bits of information can be loaded and which will
both deliver information to appropriate users and allow users to
query the system. But most important, this internal information
must be merged with all the other, more actively sought information. After all, the user doesn't tare where the gold was mined, only
that it was gold. And the gold within the company must be merged
with the gold outside.

Mining the Open Source

MINING THE OPEN SOURCE

Passive Intelligence Gathering

In 1995, the CIA held a competition. The test: who could gather the
most information on Burundi. The CIA had some of its own people
enter the competition, along with companies from the private sector. The winner, leaving the CIA's own team in the dust, was a private company called Open Source Solutions. OSS, as it's called,
produced, within twenty-four hours, a tremendous amount of information on Burundi, ranging from vital statistics to lists of scholars
who should be read and contacted. The CIA teamdead last in the
competitionbarely produced more material than that contained
in its own, quite useful but widely available, CIA World Factbook.
Apart from what this story tells us about private sector efficiency, the importance of the test was the coming-of-age of what is
called open source intelligence. As we have discussed before, open
source intelligence is material that is not secret informationneither treated as secret by the government nor as proprietary by private businesses. Open source information always dwarfs secret
information. Indeed, much of what is treated as secret by governments can be found in open source material. There are legends
about massive, covert operations that have managed to uncover
vital secretsonly to find that Time magazine had published the
same material six months before.
The advantage enjoyed by Open Source Solutions was diat it
had mastered the emerging electronic domain. The CIA, created in
114

115

an era when the primary challenge was the location of information,

used extremely complex and sophisticated techniques for capturing information. These methods were cosdy in terms of time and
money, but, prior to the emergence of the electronic domain, were
state of the art. Indeed, one of the great ironies is that the intelligence and defense communities had been instrumental in developing the electronic database systems that matured into today's
on-line services. The CIA is the largest customer for some of these
services, like LEXIS-NEXIS.
Paradoxically, having been instrumental in creating the technology, the CIA concentrated its attention on its own internal database development. These are undoubtedly powerful indeed. What
they didn't take note of was that the external databases, taking
advantage of the digitization of open source data, had become a
superb source of material. There is a tendency, both inside and outside the intelligence community, to equate open source with trivial.
Nothing could be farther from the truth. Intelligence agencies used
to spend a substantial amount of their resources hunting down and
collecting open source material. From regional Chinese newspapers to Bulgarian radio broadcasts to trade organization statistical
publications, the CIA and all other intelligence agencies spent
tremendous resources accumulating, organizing, and mining this
data.
Once in house, of course, the information was no longer considered open source. This was reasonable. Bulgarian radio broadcasts were open source only in the sense that they did not originate
as secret material. But they were effectively inaccessible to anyone
without the CIA's Foreign Broadcast Information Service, which
used listening posts to capture the broadcasts and teams of translators and analysts to translate and decipher their meaning. In effect,
the fact that collection and interpretation facilities were uniquely in
the hands of the CIA and could not be duplicated without great
expense made this information "secret." It was inaccessible to people outside the intelligence community, except as the CIA chose to
make it public.
During the 1980s, and then with increasing speed in the 1990s,
much of this "secret" material was replicated in on-line commercial

116

THE

INTELLIGENCE

EDGE

services. Masses of material were included that the CIA would not
have regarded as directly pertinent but which were of tremendous value to businesses, making these new, open source databases
priceless.

f Mining the Open Source

117

O R G A N I Z A T I O N OF E L E C T R O N I C
INFORMATION

Even a Tune magazine article digitized and databased Nas

intelligence value in excess of its printed form because you
can search and compare in seconds.
Open source electronic systems have increased the quantity of
material available while decreasing the amount of time needed to
find and acquire that material. That's how Open Source Solutions
beat the CIA. The questionfind out about Burundiwas
uniquely amenable to passive, Zone 1 searching. The private sector,
by mastering these techniques, dramatically speeds up the time
cycle for the first phase of intelligence. But it must be remembered
that not all information is available in Zone 1 and that, as you pass
into more complex realms, from the library to the gray zone, the
old-time penalties reappear. Indeed, it would be fair to say that the
CIA and other intelligence agencies will continue to dominate
Zones 4 and 5, while they are increasing their mastery of Zone 1.
The advantage Open Source Solutions had over the CIA was
undoubtedly not permanent. Regardless of how well the CIA does
its job, the growth of electronic resources creates a tremendous
opportunity for businesses looking for information.

PASSIVE, OPEN SOURCE MATERIAL.:

THE ELECTRONIC DOMAIN

The definition of the electronic domain is, of course, that the material is maintained in an electronic format. That generally means
that you need a computer to access the material. The structure of
the electronic world is roughly as shown on page 117.
The electronic domain can be divided into three realms: commercial on-line services, the Internet, and CD-ROMs. In general,
these three have more characteristics in common than differences.

The data is in digital form. That means that in all three cases, the
material is searchable using tools appropriate to each realm. Basically, in mastering the techniques needed for operating in one
realm, you are mastering the techniques needed for all.
There are thousands of on-line services, hundreds of thousands of Internet sites, and countless CD-ROMs available. The fact
is, however, that for 90 percent of the information you are going to
need, a relative handful will suffice. The trick will be to locate the
really useful ones as quickly as possible and to hone your skills for
finding the 10 percent of information that you need that is not
readily available. But basically, you can use this rule of thumb:
About 60 percent of all information you need will be electronically
available, with the best and most predictable sources of information

118

THE

INTELLIGENCE

EDGE

being, naturally, the more expensive commercial databases and

CD-ROMs. The other 40 percent will come from paper and other
sources. Of that 60 percent, most will come from no more than a
dozen sources, with five or six serving as your routine workhorse. A
small fraction will require extensive searching.
All of this, of course, begins with the computer, and we need
to say a few words about what is needed here, because the computer
is going to be your core intelligence tool, for both finding and storing information. For most American businessesand, increasingly,
European and Asian businesses as wellthe computer is already a
staple in the office, so the primary need will be to make sure that
you have good connectivity to the world outside. A wide range of
computers is acceptable, but in general, we prefer IBM clones, running Windows, just because they are so commonplace. Some online services and all CD-ROMs contain search software that is
specific to a particular platform. Virtually all are available for the
Windows-Intel or Wintel world, while an increasing number don't
bother to "port" themselves to the Mac. Thus, if you have a choice,
we would urge you to select a Wintel system, simply to maximize
your options in sites and CD-ROMs.
Your computer is going to be the workhorse of your intelligence system, as well as carrying the other burdens you put on it.
While most computers will do, getting a fairly powerful one will pay
in the long run. This is particularly true if you intend to do searches
and downloading of information while using a spreadsheet, database, or word processor. Moreover, if you are going to be doing
graphically intensive workworking with pictures, maps, drawings,
or other graphical equipmentyou are going to need both a powerful processor and plenty of disk space. This is not something you
ought to skimp on. If you already have a computer, you will reach a
point where you will want to upgrade. If you are buying a computer,
consider the following configuration:
A minimum Pentium 75 processor with at least 16
megs of RAM
A video card with at least a megabyte of memory

Mining the Open Source

119

At least a gigabyte of disk space

A 6X CD-ROM drive
A color bubblejet printer
A 28.8 baud modem
The most important aspect of your computing system from the
intelligence standpoint is your connectivityhow well you hook
into the outside world. The simplest connection to the outside
world is something called a modern, short for modulator/demodulator. A modem is simply a telephone that allows a computer to call
and connect to other telephones. A modem can be purchased
today for under $200, and Windows comes with built-in software
to allow you to operate the modem. The modem's strength is that
it is simple to use, uses existing telephone lines, and is therefore
inexpensive.
A modem allows you to hook into another computer. The simplest computers to use are those provided by large, commercial vendors, like America Online (AOL), CompuServe, or Prodigy. Apart
from information that is available on these services, like news, stock
quotes, and the like, they also provide electronic mail, or e-mail,
and connections to the Internet. These services provide their own
software to use when hooking up with them, including e-mail software and software for connecting to and browsing the net. The
modem can also be used to connect to other, higher-end, proprietary databases.
The modem connection is certainly a good, practical solution
for the gathering of information. Its greatest weakness is speedit
takes forever to load images from the Internet via services like AOL.
Somewhat faster are Internet providers that advertise in local newspapers. This doesn't save money, and it doesn't give you the other
services that on-line services do, but it does simplify the Internet
connection.
Another, higher-priced solution is some sort of dedicated,
high-speed line running directly to an Internet service provider. In
a small business, an Integrated Services Digital Number (ISDN)
line can be purchased for a few hundred dollars a month. An ISDN

120

THE

INTELLIGENCE

EDGE

line permits not only high-speed connection to the Internettwice

the speed of the fastest modembut it also provides a much larger
bandwidth than the ordinary telephone line. Think of a telephone
line as a straw. One measure of the straw is how fast it draws liquid
through it. The other measure is how large the straw is. A straw that
is twice as fast and five times as large as another draws ten times as
much liquid through in the same period of time.
If you are running an extremely small business with standalone computers, the chances are that a modem connection will
suffice for you, unless you happen to be an extremely heavy user.
However, if you run a somewhat larger company, particularly one
that has created a local area network tying several computers
together, a dedicated ISDN is the best solution. Having one means
that all members of the staff that are connected via a local area network can use the line simultaneously. This does provide one savings: You won't need to provide every computer with its own
modem and line. In the vast majority of cases, this system will allow
you to access all available databases, since most permit both direct
dial in via modem or Internet connection.

O R I G I N S OF THE INTERNET

Let's begin with what we are all most familiar: the Internet. How
good is the Internet as an information provider? The answer is it's
OK, on average. But when we say average, we're reminded of the
old statistics joke: A man is lying on the ground, his head is encased
in ice, his feet are on fire. Asked how he's doing, he answers, "on
average, OK." That's the Internet. There is some great stuff out
there. And there is some incredible trash. The amount of time
you're going to spend sorting through the Internet to find what you
need will probably cost you more than going straight to proprietary
databases. The Internet is certainly going to evolve into something
critically important, but it isn't ready to support serious intelligence
needs.
Let's begin by understanding what the Internet consists of. It
originated as a Defense Department project in the early 1980s,

I Mining the Open Source

121

I funded by the Advanced Research Project Agency. Its purpose was

to tie researchers' computers together so that they could exchange
\ information rapidly. Because most of these researchers were at universities, the practical effect of the Internet was to tie together university computers using dedicated telephone lines. Now, in those
days, most university computers were mainframes, or very large
minicomputers like the Vax, that had terminals attached to them.
Researchers would log into their accounts and have, as one of their
options, the ability to send messages over the precursors to Internet
like ARPANETlater expanded to include BITnetto each other,
as well as transmit files.
The Internet started as a toil for rocket scientists. For all its
apparent simplicity, it still helps to be a rocket scientist.
Students realized sometime during its evolution that they
could use the Internet to send personal letters to each other
transcontinentally, and what began as a serious endeavor became a
highway for ordinary and downright trivial communication. In fact,
the system very quickly became clogged with things like electronic
chain letters. The key to the ARPANET and BITnet was that they
were free. The universities provided the mainframes, terminals, and
accounts, and later underwrote the cost of connecting to nodes
major university computers that served as switching centers. Since
it was free, it wasn't rationed. By not being rationed, it grew like
crazy.
About the same time, independent of the Internet, another
phenomenon arose, centered1 around the new personal computers:
the bulletin boards (BBS). Hobbyists dedicated all or part of their
computer disk space, a modem, and a telephone line to creating
BBS that others could call in to. These BBS contained, depending
on the systems operator (SYSOP), anything from useful programs
and technical support to on-line Dungeons and Dragons or pornography. These BBS were text-based and totally anarchic. Numbers of
BBS were listed on other BBS, and folks called, sometimes long distance, to find out what the BBS had available.
A critical point emerged with something called FIDOnet.

123

THE

INTELLIGENCE

EDGE

Build around Digital Equipment's computers, FIDOnet linked individual BBS together. SYSOPs would regularly log onto other BBS
and transfer messages, download new programs, exchange bulletin
board messages, and so on. FIDOnet was followed by other BBS networks, all built around slow1,200 baudmodems and telephone
lines. The result was a delightful anarchy.
At a certain point, the BBS system and the Internet system
merged. The critical moment was the introduction, at the University of Minnesota, of something called a gopherafter the school's
mascot. A gopher was a BBS that could be accessed via the Internet.
Among other things available on the Minnesota gopher was the university library's card catalog, now digitized, and a load of other
interesting stuff. Soon every university was creating gophers, with
on-line books, programs, and everything else. Then McGill University in Montreal, Canada, introduced Archie, which searched
Gopherspace for material you might want. Later, a more powerful
search engine, Veronica, came on line, and soon the University of
Minnesota and every other sane university was considering shutting
down access.
Then came the breakthrough from text to graphics, and from
mainframes back to the old BBS idea of PC-based nodes, or web
sites. By now, commercial vendors had come in to provide connectivity and anyone who had a PC and wanted to build a web site
could. This ability increased after a company called Netscape, with
one of the more bizarre and effective business plans in history, emulated the old SYSOPs and gave away software for building graphical,
PC-based web sites, and software for accessing or "browsing" those
web sites. The net result was the world wide web.
To call the web anarchy understates the matter. The web is a
zoo, and the animals have become the keepers. The key problem is
that no one really has control of what is on the web. In fact, since it
costs very little to get on the web, everyone has his or her own web
site. A complex and extremely worthy company web site has to compete for attention with Cousin Mathilda's shrine to her lost
love. The cheerful anarchy is not nearly as cheerful when you actually need to find something. But there are nuggets there to be
found.

Mining the Open Source

123

USING THE INTERNET

Using the web for information gathering requires a browsersoftware for searching the neta search engine, and exquisite
patience. People are putting information on the web. The numberone reason: self-promotion, marketing. A great many businessoriented web sites are little more than digital brochures. The web
has become an extremely low-cost marketing tool. That means that
you need to sort through the hype, the come-on, simple errors, to
find what you are looking for. But the web does have one advantage:
Looking is free, and so is most of the information.
The key to using the web is using the web search engines. Normally, when you click on the search icon from either the Microsoft
or Netscape browser, you'll be brought to a window that allows you
to choose among various search engines.
Netscape offers access to twenty-one search engines, as well as
an add at the top of the page. These search engines are what allow
you to look for the information you need. All of them search by
matching the terms you select with the terms used in the sites that
are contained in the database. This can lead to some hilarious
results. Inputting Sahara in search of information on a desert can

c
OflwrSenlrtt:

DEVELOP INTRANET APPLICATIONS

HE NEW WAV: CL.CK iis.DOWMlo

bdascpk
Lyres
EJrlte
YJim;
=i Tfe Beclric l.fttar. SHARBWARF-COM
LxkStiurl
HolBol
ti
e&a Faar^l
ffihaWlierf?
GtT SaacrPagcs

WriCzawfer
OpenTtM Intel
Accufind
IQOliolWi* Silts
BigBagk

OWVILLAGI:
Afctrti

SenicM

WHOWHERE'
DIRECTORY Find,
Connect,
Communicate with
People and
Companies
Wortdwide

Copyright 199fi Netscape Communications Corp. Used with permission. All Rights Reserved.
This electronic file or page may not be reprinted or copied without the express written permission of Netscape.

124

THE

INTELLIGENCE

EDGE

lead to locating the web site of the Sahara Strip Club and Harry
Sahara's personal web page showing pictures of his pet lizard.
That's the weakness of search enginesthey're stupid. All they
know is the term you've selected, which they see as a string of Os and
Is. All they can do is locate identical Os and Is. Every search engine
is working to overcome this difficulty, and each does it in a somewhat different way. Search engines change so quickly that some of
this will be outdated by the time you read the book. You can find a
list of search engines, and our evaluations of them, in the appendix.
The best way to master the net and the various search engines
is by actually searching for something. Let's assume that you are
someone interested in getting involved in the pager business and
need some information on the technical aspects of pagers-just
what they do. How would you go about finding information on
the net?
The first step is to select the search engine. Let's begin with a
general search engine, like Infoseek.
When we asked Infoseek to search for information on
"pagers," what we got is shown on page 125. The first two were ads
for free pagersclearly companies hawking pager contracts. Then
there came two Motorola Bravoplus free pager offers. Then a free
pager marketing opportunity, and so on. As you can see, we didn't
find what we wanted here. Now, there were many more pages we
could look at, and buried among the come-ons there might well
have been information on pager technology useful to us in making
a decision on investment. But finding that would require a tremendous amount of time.
Notice that Infoseek, like most search engines, ranks sites by
what's called "relevance," which sounds more useful than it is. Infoseek ranks for relevance according to whether the query term is
found near the start of the document or in the tide; how many
times the term appears; and whether the term has a "high weight,"
which means that rare terms like interstitial have a higher weight
when they appear than the term pager. This shows dial relevance
means, on the whole, nothing, but it is the best that programmers
can come up witfi to screen the soup. The highest-ranking documents we found had values of 34 percent, which meant that pager

Mining the Open Source

& Home Pages

125

Free Pager! 34**

Free Pagers 34** (Sac66K)
Motorola Pagers Bravoplus Accessories Motorola Accessories Pager
34** (Sat73K)
Motorola Pager Brarophis Free Pager Offer! Features Of Bravo Plus!
34** <$a,69K)
Free pagers 34** (SiifUK)
SFS Free Pager Marketing Opportunity 34** (Silt 6 IK)
Bergman Jewelers Antique Watches Gold Diamonds Rolex Jewelry
Necklaces 34** (SnelJgK)

FREE Pagers - 30.00 Value 34** (SmJ.SK)

FREE PAGERS! 34** (Siz^llQ
Pager - GSM - Pagers - CeDular - Phone - - Telephone Directory 34**
MobileMedia

(Bize73K)

Make Money Giving Away Free Pagers Work at home Opportunity

ui

^j

Reprinted by permission. Infoseek, Ultrasmart, Ultraseek, iSeek, Quickscek, Imageseek,

Ultrashop, "proof of intelligent life on the net" anil the Infoseek logos are trademarks of
Infoseek Corporation which may he registered in certain jurisdictions. Other trademarks
shown are trademarks of their respective owners.

was a common word, appearing not all that frequently in the text
which means nothing. Relevance on the net is a case of false
precision.
On the Internet, nothing is more irrelevant than relevance.

Obviously, this first search yielded little. We shifted our search

to an engine called Sawysearch. This is what is called a metasearch
engine, which searches about twenty search engines simultaneouslyin other words, rather than going to one search engine after
another, you can just have the metasearch engine do it all. Now, you
might think that by searching twenty infoseeks simultaneously, the
data dump would increase. But Sawysearch and the other metasearch engines do two extremely valuable things. First, they merge
together similar categories. For example, the multiple hits on "free
pagers" that swamped our first search are merged into one single
entity. You can choose to click on that line and wade through the
free offers if you want. But you can also ignore them.
The second thing that the metasearch engines all allow is
something called a Boolean search. These are extremely important

126

THE

INTELLIGENCE

EDGE

and not quite as boring as they sound. Basic search engines allow
you to search for key words. Be sure to check the tips given when
using a particular search engine to see if it is case sensitivesome
are, some aren't, and it could make a difference. So, if you were to
input John Kennedy, every site where the terms John and Kennedy
showed up would be listednot very useful since you'd probably
find ads for Bill Kennedy the John cleaner in a non-case-sensitive
engine if you didn't specify. Boolean searches allow you to gain
greater control. You can specify, in the best engines, the distance
from each other that two words have to beso that you are only
interested in John Kennedy when they occur next to each other.
They can specify that you are interested in John Kennedy and Lee
Harvey Oswald, and only those sites containing both names will
appearor you can say that you are not interested in any site containing both names, in which case John Kennedy will only turn up
when Oswald isn't mentioned.
Between merging output and Boolean searches, the metasearch engines are much more powerful tools. What we got when
we input pagers this time is shown below. The first thing that
shows up is for an electronic mosquito repellent! We spent precious
minutes looking at that because...well, you've just got to, right?
And that's one of the real problems on the net. The ratio of inter-

Mining the Open Source

127

esting but useless information is so high that you are inevitably

going to stop to look at the weird things that turn up. But some
other things turn up as welleach with the search engine that
found them listed to their right.
We clicked on pager sales and service, which gave us, among
other things, the home page of a pager company in Vermont (see
below). Rare among business "brochures," this page gave us at
least some minimally useful information on pagers, including what
they do and linking us to manufacturers. Of course, while more useful than other sites, we still haven't learned much that we didn't
know beforeor to put it differently, if we didn't know this stuff
already, we had no right to be interested in the industry.
Anyway, as you can see, we were getting nowhere fast on the
net. We turned our attention to another, somewhat less well known
part of the Internet, called newsgroups. Newsgroups predate the
world wide web graphical home pages, evolving out of the old bulletin boards. Basically, these are places to post messages. There are
currently about Fifteen thousand newsgroups, with the number
growing continually, each located on someone's computer somewhere, and each accessible through various Internet toolsand
each dedicated to some topic, from bestiality to quantum mechanics to everything in between. The Usenet Calculator estimates over

V&M Distributes Ere seats Electroiac Mosquito Sefr-Ja (WebCrawler)

UBS pager wifl sound an alert or silent nobficatton eacfi torn a message is received and ihe phone
nurabef Irftbythe taller will be displayed <* the pager This imiwffl store between 6-16
messages in it's memory, some tone stamp each page, and each m3 also a&>w you to sead
codes

wireless (WebCrawler)
_Mail (WsbCrawler)
Motorola m Seutschlan<ivAktrBj-aten i 93? (WebCrawler)

Pjgef Sales & Service (WebCrawler)

g&Bjifrwce^Ercj^PagerJft'

. (WebCrawfer)

Bravo

Faeer Autumaaon (Yahoo)

Iri'erNet Faeer (Yahoo)
Weggia Wjtchei j'ager Semce (Yahoo)

.14**
Reprinted wiih permission of SavySearch.

Reprinted willi permission of Coniact Communications.

128

THE

INTELLIGENCE

EDGE

eight hundred megabytes, or the equivalent of eight hundred hefty,

four-hundred-page novels, and is posted to Usenet newsgroups
every single day of die year. This amount has, in recent years, heen
increasing at the rate of 100 percent (i.e., x2) every year. That's a
lot. Chances are, what you're interested in is discussed in excruciating detail somewhere in Usenet.
When the president's press secretary rails about conspiracy
theories being published on the weh, when Pierre Salinger talks
about an intelligence document he found on the web, when we talk
about neo-Nazis using the web, what we are really talking about are
the newsgroups. These can be searched in the same way, and with
similar engines and tools, as the web sites we started looking at.
We used a search engine called Dejanews. Looking and working
very much like other engines, it searches through this mass of
rubble.
Well, there may have been something useful here (see page
129), but we couldn't find it. What we did find was a mass of pagers
for sale and requests for programs that would crack pager control
systems. There is an awful lot of junk here. But these discussion
groups do provide one thing of value: They point to people who
might be interested in the same thing you are and give e-mail
addresses. All of these hackers trying to find pager program crackers probably know an awful lot about pager technology. If you're
really up against it on a project, it's not a bad place to tryassuming you're careful about who you're dealing with and how much
you give away.
We pushed the search one step farther. We refined the search
from "pager" to "paging industry," using another metasearch
engine called Motherlode Insane Search, which in our experience
provides some of the better results by ranking things a little more
accurately (see page 130). We clicked on the second line, Industry
Links, provided by Lycos, and that took us to another page, and
another. What was finally yielded is shown on page 131. This contained an extraordinary collection of information on the technology of pagers by someone who used to work for Motorola and now
worked for a company called Real Time Strategies. We also found,
from VHF Communications, a brief history of wireless communications, all culminating in the pager. From Pagetrac, we found an

129

Mining the Open Source

Date

1. 97/02/06
2. 97/02/06
3. 9 7 / 0 2 / 0 6
1. 97/02/05
S. S7/02/05
6. 91/02/05
7. 97/02/04
B. 9 7 / 0 2 / 0 4
9. 97/02/04
10. 97/02/04
11. 97/02/04
12. 37/02/04
13. 37/02/04
11. 97/02/01
15. 97/02/01
16. 9 7 / 0 2 / 0 4

inis*

Set
027
027
027
027
026
026
026
026
026
026
026
026
026
026
026
026

Subject

Newsgroup

Author

Jt

TB: Pager Interface (PI tnn. COB. pager

rbarnett^tcavi
P1GEP
pl.ogldizenia.sprze auc@ ib ib . uav .
camp . 13* . ita- as400 . cer793vSur3a.:
AS/100 Pager support
Ignore: Pe : Jl Pager Busl akr.freenet
redflredpoll.K '
Pager Gateway
cmqi.gr oupware.iotu JMPLB7Becrodi. tnn . count, pager
tirinspetes8ao :,;
A f r e e 530 pager
Pager advice -* numeric o CQc . culture . singapo lukyni@hormaLl "'
Motorola pager for sale tiui . CUM- pager
"H. Etchaniend^*

Weed Pager program crack alt . 2600 . crackz

Need pager program crack alt . crack*
Need Pager proaraitl crack alt .crackers
t.i^-tHAi, [ ksijiw.a?
han.inisc.Bisc
NEED A PAGEP, SAVE
alt . cellular
FREE
SBO pager Isc . indtutry . elect
pager
pl.ogloiEenia-sprze
Paetec Control
coae.lang. basic, via

voyager IBrny.ni,
Vo vager Igniv. nr
Voyager Igmv . ni
"Perioua" <da '
"acs" <acspli\ .

twinapetespao
"inea" ^ineEp 1
Peter Nguven
n.Tne,=,Y -' - B^:->
-i-
^~-"----.i

Reprinted with permission of Deja News, Inc.

offer to sellfor $1,200a study on Consumer Trends in "paging."

And we found most of the corporate web sites.
In the end, we found a great deal of information, all of it for
freesave the one study being offered that we didn't have to buy.
But the free information came at a high cost. It took us about six
hours to carry out the search and about two hours to sort and organize the material. The search required someone with enough
sophistication to distinguish the significant from the insignificant,
so it was far from minimum-wage work. A day to nail down the information diat we found means that it cost us a few hundred dollars.
Now that wouldn't be bad if we had found everything we needed,
but what we found was quite spottycertainly not enough information to make any sort of decision on whether or not to invest in
an industry or a company.
And that's the problem with the Internet. There are certainly
valuable things to be found, but it is unlikely that you will derive a
comprehensive situational awarenessone sufficiendy robust to
allow you to make a significant decision. Nor can you predict what
class of material you'll find on the web. The web's chaotic nature
means that, from industry to industry and from time to time, you literally might find anythingor nothing. And diat nothing can cost
quite a bit, considering the amount of time it takes and how much
that time costs.

130

THE

INTELLIGENCE

EDGE

Worid Wide Web

VHF ComrnmucatiQfls. Inc - A Bnef History QFWireless (Webcrawler)

' feiduslr? LittN (Lycos')
Minuses Global Maricets - Annual Report \99f; - Extract from ihe 1996 Memtec Annual
Report showineMerntec's Gobal Markets (ALIWEB)
sagdaa var_eli?ss iMrnejuracaiiogs. facilities issues paper (Webcrowler)
Ajsoi:iahqii_Pubkcatijgns (Lycos)
Tgnsit Liduslry Sofhvafg - Information on CSI's capabilBes for the Transit Industry
including computer- aided dispatch, commumc aliens and schedule management (ALIWEB)
Sp_eciai Report; {Webcrawler)
A_jndtpeadarJ Teleccm Cnnsijitants esteri?!ve sitejsr '96_E (Lycos)
HgtElbuyetj;om - HotelBuyer.com is deagnedto be an electronic meetmg place where
Hospitality Industry related business people can come together to Snd leads for and offer
products to others in Ihe Hotel Industry. We list vendors who stU to the Hospitality Industry.

Mining the Open Source

131

':

Reprinted with permission of USWeb.

The Internet is free only if your time's value is zero.

That's why, for us, the web is rarely the first stop. Unless your
time doesn't cost much, or the project is not that significant, we
tend to shun the webuntil the end, when we're just looking
around for things we might have missed. In our experience, a high
degree of confidence in knowing where to find what you need more
than makes up for the money that must be spent on that confidence. One day the web may be a good place for intelligence, but
right now, we prefer the commercial, on-line services.

O N - L I N E SERVICES

On-line services are peculiar creatures. They are not, themselves,

information providers. Rather, they are middlemen. They provide
outlets for information producers. They provide the equipment
the computers and phone lines and so onas well as the market.
In return, they take a fat percentage off the top. Sometimes they
buy a single source, like a newspaper, and sell access to it. Sometimes they buy entire databases, like the Hoover Directory of U.S.
Businesses. They also provide the search engine, which, in many

Reprinted with permission of Bradley F. Dye.

cases, allows you to search for information across several data

sources simultaneously. Rather than having to search each newspaper individually, the search engine allows you to search all of them
simultaneously. That simultaneous search capability represents
a fundamental increase in analytic power. It is also something you
pay for.
Anyone wanting to do business intelligence must master the
tools of the trade. The basic tools of this trade are the on-line services. There aren't that many good general ones, but each has its
strengths and weaknesses and the serious practitioner must know
when to use each.
One of the most popular and powerful databases is LEXISNEXIS, whose origins we have already discussed. LEXIS-NEXIS provides a tremendous amount of resources. When you log on to
LEXIS-NEXIS, you are provided with a host of options (seepage 132).
This is the first menu page on LEXIS-NEXIS for the NEXIS side.
The LEXIS side is primarily of interest to lawyers and need not interest us here. Note that NEXIS is divided into what are called libraries.
The libraries divide between general news libraries and specialized
topics ranging from industries to various geographical locations.
There is a substantial overlap between the sources in each library. A
particular publication might appear in the News Library, as well as

THE

130

INTELLIGENCE

KftGE

Mining the Open Source

131

World Wide Web

'

VHF Communications, Inc - A Brief History Of Wireless (Wsbcrmtler)

Industry Links (Lycos)
Merntecs Global Markets - Annual Report 1996 - Extract from the 1996 Memtec Annual
Report showing Memtec's Global Markets (ALIWBB)
saadaa wireless cymmuracalions facitoes issues paper (Webcrawler)
Association Publications (lycos)
TransjtIridijstry_gcBware - Information on CSI's capabilities for the Transit Industry
including computer-aided dispatch, communications and schedule management (ALIWEB)
Special Reports (Webcrawler)
A Independarjt_Telec<ini Consultants extensivesi!; for '96 p (Lycos)
Hotelbuyer com - HotelBuyer com is designed to be an electronic meeting place where
Hospitality Industry related business people can come together to find leads for and offer
product! to others in the Hotel Industry. We list vendors who sell to the Hospitality Industry

Reprinted with permission of USWcb.

The Internet is free only if your time's value is zero.

That's why, for us, the web is rarely the first stop. Unless your
time doesn't cost much, or the project is not that significant, we
[end to shun the webuntil the end, when we're just looking
around for things we might have missed. In our experience, a high
degree of confidence in knowing where to find what you need more
than makes up for the money that must be spent on that confidence. One day the web may be a good place for intelligence, but
right now, we prefer the commercial, on-line services.

ON-LINE

SERVICES

On-line services are peculiar creatures. They are not, themselves,

information providers. Rather, they are middlemen. They provide
outlets for information producers. They provide the equipment
the computers and phone lines and so onas well as the market.
In return, they take a fat percentage off the top. Sometimes they
buy a single source, like a newspaper, and sell access to it. Sometimes they buy entire databases, like the Hoover Directory of U.S.
Businesses. They also provide the search engine, which, in many

* A collection of information about Radio paging

Reprinted with permission of Bradley F. Dye.

cases, allows you to search for information across several data

sources simultaneously. Rather than having to search each newspaper individually, the search engine allows you to search all of them
simultaneously. That simultaneous search capability represents
a fundamental increase in analytic power. It is also something you
pay for.
Anyone wanting to do business intelligence must master the
tools of the trade. The basic tools of this trade are the on-line services. There aren't that many good general ones, but each has its
strengths and weaknesses and the serious practitioner must know
when to use each.
One of the most popular and powerful databases is LEXISNEXIS, whose origins we have already discussed. LEXIS-NEXIS provides a tremendous amount of resources. When you log on to
LEXIS-NEXIS, you are provided with a host of options (seepage 132).
This is the first menu page on LEXIS-NEXIS for the NEXIS side.
The LEXIS side is primarily of interest to lawyers and need not interest us here. Note that NEXIS is divided into what are called libraries.
The libraries divide between general news libraries and specialized
topics ranging from industries to various geographical locations.
There is a substantial overlap between the sources in each library. A
particular publication might appear in the News Library, as well as

132

THE

I N T E L L I G E N C E

EDGE

Mining the Open Source

133

lea^.e t n i r n the NHFlb <.only onej of the library you iirtnt to y s e _

To see (lie
lescription of a particular library, type its page number and then press the
INTER keyNftnE

PG HUME

PG NRP1E

PC NOME

PG NBHE

Types
General
News - - Legal NEWS
1 LEGNEU 1
REGMUS 1 LEXPfll 4
TOPHUS 1
Financial
flCCTG 2
CCH1PMV
2
NfiBRS
2

BflNKNG 33
BUSF1N 33
CMPCOM 3
ENEHGV 33
ENTFRT
2
ENU1RN
2
INSUNE 3
MftRKET
1
PEOPLE 2
SPORTS 2
IRftNS
3
Reference
DUKHEF
3

PG NAME

Topics

PG DMArtE

- II nnft. ' I}
ASIflPC
Medical
DSIflPC
GENHED
GENMED 5 CANADA
EPIBftEE 5 DUTCH
MEDLME
MEDLNE 5 EUROPE
GERMflN
P o l i t i c a l P1DEAFR
ftPOLIT 6
MEXICO
CMPGM
6
NSflHER
EXEC
6 UK
LEG IK
& UKCtIRR
WORLD

ftssist:.

GUIDE
TERMS
PHftCT

PG
-7
5
B
7
77
7
7
B
7
5
5
7
8
8
8

-;or further explanation, pi-ess the H key (for HELP> and then the ENTER k e y .

This chart and those on pages 133 and 134 are reprinted with the permission of LLXIS-NFXIS,
a division of Keed Elscvier Inc. LEXIS and NF.XIS are registered trademarks of Reed Elseiier
Pro]>er!ies Inc.

in World, UK, People, and so on. You cannot search all of NEXIS
simultaneously. You must, rather, choose a particular library.
Each library is subdivided into individual files. These files can
be searched individually or collectivelyalthough sometimes some
files are excluded. Assume that we were interested in information
on particular companies. We would go into the Company Library,
which is reproduced on page 133. This gives you a series of options
from which to choose, including everything from bankruptcy
reports to mergers and acquisitions. You can also choose to search
for everything simultaneously.
Let's assume that we were to do a search on our pager problem in NEXIS. We would go, for example, into Alnews and do a general search on pagers. We were informed that over a thousand
articles would be available with the term pager. We therefore narrowed the search a bit to "pager manufacturer," which still gave us
485 stories (see page 134).
Notice the mix of press agency stories and industry media. It
now takes time and patience to sort through these stories, which are
full-text, and find what you are looking for. It is important to understand a little of how this works, because watching a computer sort
through literally millions of individual articles looking for the word
or combination of words you've put in, and then calling up the
story in well under a minute, is a pretty remarkable thingand the

'lease ENTER the NBHE of tire file you want to search.

To see a dascriptinn
,f a file type its page nunher and press Che ENTER key.
FILES - FBGE 1 of 16 <HEXT PflGE for additional f i l e s )
NftrlE
PG PESCRIP
NOME
PG DESCJtIP
HUME
PG PESCRIP
C O U P O N ?
L I B R ft R Hens ft
Business -Gonpany Reports By Type
SEC Filings
N(S1J2
* Business
F I L I N G 1 SEC Full-Text flLLNHS 1 Comnany News*
BDS
1 Bankruptcy Rpts
ftLLfiBS
IS Company flhstr
EXECDR 1 BUS Bio Reports* EDGflRP 1 EdgarPlus
SECOL
1 Sec Online
BUSOPP IE I n t l Bus Opps
1 US Co Rpts
USCO
ftCCESS 1 flccess Disclo* - flssists
1NTLCO 1 r n t l Co Rpts
- U . S . Conipanies
G U I D E Ifi Descriptions"
flLLMB
1 rtfi Reports
IISPHIU
2 US Pi-iuate Cos
SFC
1 SEC Filings
USPUB
2 US Public Cos
STKRPT 1 Stock Renoi-ti
HESS NEXT FflGE FOR flDDITIOMflL SOURCES.
Files narked nay not be combined.

VEXIS Survirr

core phenomenon of our time. In effect, every word in every article

and even,' database is entered into a master database. Each word, in
turn, has entered after it identifications that point to articles in
which that term appeared. Every time an article is entered into the
database, it is broken up into words. Where a word already exists,
an entry is made stating that such and such an article also contains
that word. If the word doesn't exist, it is created with the appropriate notation. So, when we enter the word pager, the computer
instantly goes to that alphabetical entry, finds all the articles that
have the word in the title, and lets the user know that the article
exists and offers to call it up for him.
All of this is, of course, expensive. Our search on pagers cost
about $725 on a per-use basis. Now, you can get a flat rate on LEXISNEXIS for $4,250 a monthplus print costs, which run about $2 per
document. Obviously, we try to control our LEXIS-NEXIS costs. It is
basically a trade-off between labor costs and search costs.
An alternative to LEXIS-NEXIS is Dow Jones news service. This
comes in two flavors: a web-based service called DowVision, which
focuses on feeding real-time news to you, and the dial-in service
Dow Jones News Retrieval, which is more of an archival service,
although there is a great deal of overlap between them. Dow Jones
tends to have more detailed financial information than LEXISNEXIS, but the breadth of its global news retrieval is not up to LEXISNEXIS standards. When we searched for pagers on Dow Jones, we

132

THE

ol

INTELLIGENCE

EDGE

Mining the Open Source

133

.. t j
, ... ...IT- L j.ur.try <r"'i uf>"> to <i->e To -.ee
d p.mti' t i |flih I ibiMi-y, ' vpe il -, P-"t'' nupiJier -tnd ' F^r> prei .

Nfi"i

1 I.J.GNFU
r LFXPflT
1

Pf: Nflhi.

H; wnriL
RHhKNI,
BDSUN

1
1

CMPCOH

LLTSRFF

'3'

i'b HUME
3
j

FNLKliY
LKIFRI

.1
2

PFOPLF

rcJ HftriL

ih^ DNnriF

- .1
GFNMED b

flSIBPC

FC
'!

flFDLNF

LLISOPL
(iFllflfIN
!'.J i .t ;. i ! riDFA}-H

spoLii
OIPCIN

t
K

FXLCLFCUS

6
t,

7
7
7

nm co 5
NsnntR 7

UK
UKCUfiR
IJORI.I)

i c FNJTR tli-- HfifM1 L>| I lie file yrm u.int to s e a j ' c h Ju see a d ^ " < - > - i ,
.1 .1 1 r l ' - r ypf! lUs fJ^ge nutiln'r .nul press rh.- T N I L K k e y .
Fll.Pi
I'HCL 1 of 16 <NFXI t'flliL t o r a d d i t i o n a l f i l e i >
Mflll
I 1 '! I'MiuKlf
Nfihl'
I'G DEECRIP
NftMF
PC, DtKCRIP

S
b
7

1
1
I
I
[

Gil I lit ' M

i tun;;

1 1"

Itiinkj'uptc y ft|it .
BUS R i . i H o y o r t s *
US Co Rpts
I n r l Cn Mjjts
naB Retort i

i fire r i im

IKhl'l

1 S t o c k Rr^i

V
I. I !' i: ir I! '/
'< l_Linu=
1 RFC full-Ie.t
1 tdgarPl'i'.
I Si-i- Online
1 Aere*^ IJK.ulo.'.
.i.-ji..,!!.
-2 US P v i u a t e Cos
1 Uis Public Cui

NFXT I'fKJl tUFt ftDDITIONflL SOLRCFS.

This chart anti those on pages 133 and 134 are reprinted with the permission of LEXIS-NEXIS,
a division of Reed Elsevier Inc. LEXIS and NEXIS are registered trademarks of Reed Elscvier
Properties IDC.

in World, UK, People, and so on. You cannot search all of NEXIS
simultaneously. You must, rather, choose a particular library.
Each library is subdivided into individual files. These files can
be searched individually or collectivelyalthough sometimes some
files are excluded. Assume that we were interested in information
on particular companies. We would go into the Company Library,
which is reproduced on page 133. This gives you a series of options
from which to choose, including everything from bankruptcy
reports to mergers and acquisitions. You can also choose to search
for everything simultaneously.
Let's assume that we were to do a search on our pager problem in NEXIS. We would go, for example, into Alnews and do a general search on pagers. We were informed that over a thousand
articles would be available with the term pager. We therefore narrowed the search a bit to "pager manufacturer," which still gave us
485 stories (see page 134).
Notice the mix of press agency stories and industry media. It
now takes time and patience to sort through these stories, which are
full-text, and find what you are looking for. It is important to understand a little of how this works, because watching a computer sort
through literally millions of individual articles looking for the word
or combination of words you've put in, and then calling up the
story in well under a minute, is a pretty remarkable thingand the

S... ^ K i . ' i i i h -
Sl.l.NUi;
1 Conpani; Nuns
ftLLflBS f, C^jiipari) flh^t
RllSUPf IS Intl Hiri. Opt
n ..I.:
CUILL
16 DC- i j-iplions

Tiles narked

NEXIS Service

core phenomenon of our time. In effect, every word in every article

and every database is entered into a master database. Each word, in
turn, has entered after it identifications that point to articles in
which that term appeared. Every time an article is entered into the
database, it is broken up into words. Where a word already exists,
an entry is made stating that such and such an article also contains
diat word. If the word doesn't exist, it is created with the appropriate notation. So, when we enter the word pager, the computer
instantly goes to that alphabetical entry, finds all the articles that
have the word in the tide, and lets die user know that the article
exists and offers to call it up for him.
All of this is, of course, expensive. Our search on pagers cost
about $725 on a per-use basis. Now, you can get a flat rate on LEXISNEXIS for $4,250 a monthplus print costs, which run about $2 per
document. Ohviously, we try to control our LEXIS-NRXIS costs. It is
basically a trade-off between labor costs and search costs.
An alternative to LEXIS-NEXIS is Dow Jones news service. This
comes in two flavors: a web-based service called DowVision, which
focuses on feeding real-time news to you, and the dial-in service
Dow Jones News Retrieval, which is more of an archival service,
although there is a great deal of overlap between them. Dow Jones
tends to have more detailed financial information than LEXISNEXIS, but the breadth of its global news retrieval is not up to I.KXISNEXIS standards. When we searched for pagers on Dow Jones, we

134

THE

INTELLIGENCE

EDGE

Mining the Open Source

135

LEUEL 1 - 485 STORIES

.. The associated Press, February 3, 1997, Monday, DC cycle. Domestic Heus, 949
lords, BTUWTfl
HP Online, January 31, 1997; Friday, Financial pages, 2654 words,
inancial Heus at 9:18 a.m. ESI Friday, Jan. 31, 19S7

HP

I.
Co input ergram International, January 31, 1997, No. 3990. 107 nurds. BANKS
I I B N D BY BANKRUPT PAGER rlOBILEMEDIB. IflC 06231375
I.
Neusbi'tes, January Ji. 1997, 610 words. MohileMeilia Files For Chapter 11
"rotection 01/31/97, IflC 06231840
I.
f l . j i ' M i r France Fresse, January 38, 1997, Financial pages, 208 words,
ipproues foreign projects worth 57 million dollars, NEW DELHI, Jan 30

India

- ftP O n l i n e , January 30, 1997; Thursday, Financial pages, bb7 uords,

ItaCistic Freedom d i n s Over NBA, LBRRV NEUMEISTER, HEU VORK
ftP Online, January 30, 1997; Thursday, Sports, 867 words,
tatisties Lawsuit, LflRRV NEUMEISTEB, NEW VOHK

NBA Loses

NFXTS Servire

found overlap between many of the more recent articles, but found
that LEXIS had a stronger historical base. Dow Jones is a good
choice if your needs are focused on fairly recent materialand it is
substantially cheaper to use than NEXIS-LEXIS.
There are alternatives to LEXIS-NEXIS and Dow Jones that are
powerful and cheaper. Eye Qis a smaller but still powerful on-line
sendee that focuses heavily on business needs. For a fee of $39 a
monthplus $3 per article retrievedEye Q provides a scaleddown version. Eye Q's business menu (shown on page 135) gives
you a sense.
Notice the simplified user interface. While generally desirable,
we find that simpler interfaces frequently gloss over the fact that the
search engine is less powerful or the data contained on-line is less
extensive. Powerful search engines and large, yet focused, libraries
seem to require complex interfaces.
Wben we ran our search on pagers, we found that we did not
get the kind of coverage that we got I'rom NEXIS, although we did
get decent information at a fraction of the price. One strength in
Eye Oj market reports that were not available on NEXIS. We spent
about $125 to $150 looking for pager information. Our results are
shown on page 136.
Among the cheapest sources and an interesting approach to
initial screening is E Library, which is primarily a newspaper and

Reprinted with pennission nf UM1.

periodical archive going back to about 1990. It allows full Boolean

searching and is divided into libraries (see page 137). Its great
advantage is price: $9.95 a month, unlimited access, unlimited
downloads. At that price, we think it's a powerful tool indeed. Wben
we ran our pager search there, we found only about one fifth of
what NEXIS gave us, and perhaps one tenth of what we did have
could not be found in NEXIS. But it did provide us with some basic,
useful coverage.
TAG Insite consists of six different databases: ranging from the
lop-of-the-line Market Insite, which covers a wide range of current
publications, news wires, press releases, and so on, down to the
highly focused Health and Wellness Insite. Market Insite allows you
unlimited access for a yearfor a hefty $5,995. For that you can
access data from about two years. One of the nice aspects of IAC is
thai, it contains brokerage firm reports and summaries of companies that are usually not available on LEXIS-NEXIS or Dow Jones. It
does not have industry periodicalsa substantial lack. IAC allows
you to access information much more easily than the others, and
includes hypertext links that arc not available elsewhere. Its flat rate

136
THE

I N T E L L I G E N C E

EDGE

Mining the Open Source

137

10'1& documents found

WHO

w^*"-

'*--

Research

WHAT psjBr
WHEN' This rr and Last ft,

Real Remand NOW!

WHEHE Global

paging industry

p**
MLWK

w.

in

Q4/OT/96 SERVICE COSTS MOWTHLYFEE

PAPLRG
ON CALL TEENS TURN BEEPERS. CELL PAPLRG
Pa^eis sll coni(jeliig n ceMicne waW BUSJOU

2
SLMO 02/18/36
3
DLBJ
11/15/96
4
HOEJ
01/12/36
5 DAL
02/11/96
6 AAS 01/06/96

Instant flttl Trial!

SLTH
VADT
JPOE

Enter a question or somt keywords Select source types Click Go I

Vitw Selictinnt

Try it now and receive a free 30 day tnal Click here for details

BUSJRL
RINGING UP TEEN SALES CeUaf.
PV5PLRG
Pajei Pal is the latest in gunjfcigei
PAPMED
04/19/96 F*Th4tBeap!TagetheiStayitBn" FAPSML
08/01/96 Inda Pagng
" MAGOTH
09/02/36
idTrs^PJvKtanss nan
PAPSML

" Magazines

P Newspapers & Newswires

Hundreds of fgll-^eKt

Deliver, Srvic
Dictionanr wd

Maps

P TV & Radio Transcripts

Tf If vision transcripts

Hundreds available
. y " --'!'--

11
12
13
4

CTMA 05/16/ffi
AWEK 05/04/96
NWCP 01/1S/97
TBJP,
09/2?/36

lisues.PsgeirrwketitiholdspolenBal
BUSJOU
COMMUNICATIONS- BuWng a Belter
MAGNWS
MakingTheCaiWjthTwo-WavPsgng " 'TRODE
Paoer: Beyond mere beeping
BUSJOU

3
2
5
2

PrJvBte E ye

Conacton Question J

Books

P Pictures

Over 2.DOO works

Ovn 23.000 Photos

Cloie
View Options:

FULL

Reprinted with permission of Iiitonauiios, Inc.

Help

UM1

sounds steep, but compared to LEXIS-NEXIS, it's a bargain. It gave us

many of the recent stories on pagers that we got in LEXIS-NEXIS, but
not all of them.
Profound is accessible as a dial-in service. A base package is
offered tor a fee of $198 a month. But many of the files you'll want
cost extra. Profound is heavily focused on company and market
news. Il gives good graphics and therefore provides excellent
charts, but it has relatively little news outside of this: New technology, which is what we are interested in, is not emphasized, except
where it impacts on companies. This is a great place for people playing the stock market or for companies needing quantified information, but it won't do for more general searches, and the price
can add up.
There arc other services. Proquest is an extremely expensive,
relatively sparse database thai can be avoided. In fairness, it did give
us a great article on the European pager market, but we finally
found ihat elsewhere and cheaper. FT Profile is similar to LEXJSNKXIS in price but not in breadth and is both difficult to access from
the United Slates and lacks (although this is said to be changing)
powerful search tools. Others spring up and drop out regularly. But
starting a database sendee takes time. The established databases

have a tremendous advantage over new services: They've already

stockpiled a huge database. So, while others will be entering the
fray in niche markets, we expect fewer rather than more business
databases.
There is another class of on-line service that you should bear
in mind when looking for informationwhat we call the general,
mass-market on-line service, like America Online and ComptiServe.
Both are intended for multiple functions, including providing
e-mail and Internet access, general education and entertainment,
travel planning, and so on. While each contains business information, we Find that CompuServe is far and away the more powerful
business service. CompuServe provides a service called Executive
News Service (ENS) for $10 over the regular monthly subscription
of $24.95 plus $1.95 an hour over twenty hours. ENS is an electronic
clipping service. You can inpuf a word or a company and anything
appearing in up to twenty-four news services will be delivered
to you.
Perhaps more important are the additional-fee business services available via CompuServe (see page 138). While some of these
services, such as Dun and Bradstreet Reports, can cost a bit, others,
like Hoover, cost liltle or nothing. We find ourselves regularly using
specific services on CompuServe, simply because they are not easily
available elsewhere.

138

THE

INTELLIGENCE

Business

EDGE

marketing/
sales
Company/Industry
ources

age ad tor Fortune magazine lhal

votid have diawn allertiorj to *s recent

Dun S Biaditree:
Onfine

Deniograpftic
Research

Hoover's
Company Si tec I pry

Inte manorial Company

siflii^rrfliEhfieh<ilderwBfl^aDc,V'JDrre.
4 Co. i/rt^rqecledbyihEWelfSbee'Jcrum
The iragKine wil pm ih* bannecf ad BI
TheNewVoikTime
Foruuu

\ Heir ? ^

Wlll-3

Hen

Mail

Con't get to a computer?

No Problem!

you can listen to your e-mail over the phone

Reprinted with permission of CompuServe.

AOL is attractive because it carries a great deal of material

from Dow Jones. Some of this is identical to the material contained
in Dow Jones's own business databases but is included in AOL's base
price. A CompuServe and AOL account will actually give you a great
deal of current business information. What you won't have is a great
deal of archival informationstuff' from five years ago. You also
won't have powerful search tools. But having the two accounts will
cut down the research you have to do on more expensive databases.
But neither AOL nor CompuServe yielded very much on pagers.
Once we had companies we were interested in, like Motorola, both
could provide us with some basic, useliil information for very little
time or money.
A word about MSNetwork from Microsoft. It is really geared to
family and home. It has some personal investment information but
very little for business. This is likely to change, but at this point, we
wouldn't class this as a business investment.
Now let's pause for three other on-line sources that are heavily focused on technology: Dialog, National Technical Information
Service (NTIS), and Defense Technical Information Center. The
last two are actually run by the federal government while the former, discussed before, is currently owned by Knight-Ridden These
are great places to find technical information on anything, but they
are all costly. Not surprisingly, the two government databases are
the hardest to access, the hardest to search, and quite expensive for

Mining the Open Source

139

information that is supposed to be free. Dialog is notoriously hard

to use and expensive as well. But we found a wealth of information
on paging technology, much of which was so technical we weren't
capable of following it.
Finally, let's remember libraries. A great on-line source is Colorado Alliance of Research Libraries (CARL), run by the University
of Colorado. This is accessible over the Internet (via Telnet) and
archives abstracts of recent articles in popular and not-so-popular
journals. The search is free. They will also fax articles to you for
about $10 a throw, but you can usually pick up the same article at a
local library or university for free. CARL is not useful for very current material, but it's great for historical articles. And this brings us
to the final point: libraries. In the end, you've got to go to them. But
their catalogs will come to youand you'll find an awful lot there.
The appendix gives you a list of the better on-line library catalogs
almost all of them free.
In the end, we found an article on pagers on LEXIS-NEXIS from
the St. Petersburg Times. The article was about a local company that
had made a fortune in pager repairs. The article pointed out that
pagers are made to last about two years. After that, they need to be
repaired. This one litde article led us to Dialog, NTIS, and then to
the library for an answer to the question. If you want to get into the
pager business, think about pager repair and refurbishing and stay
away from sales.

CONCLUSION:

REMEMBERING

THE T R I E D A N D TRUE'

In the end, the most extensive and cost-effective sources of information are the vast research libraries that until recently represented the sole information archive of our society. These
librariesand the books and journals they containremain and
will remain the foundation of our society's knowledge base. Paper
remains a powerful storage medium. It is durable and user-friendly.
The great weakness of paper is accessibility, and that is a problem
difficult to overcome.
Electronic sources can be searched by the word. Paper sources

140

THE

INTELLIGENCE

EDGE

cannot. It is impossible to know what is contained in a book or arlicle without reading it. Titles may or may not provide the needed
information. Subject headers hint content, but they are rarely
exhaustive and can even be inaccurate. Even author-prepared
abstracts of articles may not reveal information in which the author
was not interested but in which you might be. That is the crucial
problem with paper. Regardless of how powerful (he cataloging system might be, it only catalogs information about publications and
not the publications themselves.
There are plans to change this. Project Gutenberg is an academic project designed to digitize great books. The Library of Congress has a program under way to digitize its collection. All of these
will ultimately fail. Even today, far more information is being produced on paper than is being produced electronically, and the
transformation of the historical archive into a digital form will outstrip society's resources for a long time to come.
Businesses must, in their passive intelligence, master libraries
the way they are. Fortunately, the cataloging system for books
and periodicals has changed tremendously over the past twenty

EI*
1 O'Sliea, Dan
Paging becomes fl Two-Way Street.

2
Going Global or Staying Closer to Home.
3

(Telephony. 11/2B/

(Communications.

(Communications.
Outlook on the Paging Industry: 1995 Suruey Results

99/B1/

(Lan magazine.

99/81/9

(Communications.
Paging industry Surweji Operators Speak Out.

09/81/9

Internet B Work.

(Communications.
Wireless Forecast.
7

1 a/01/9

. . . 9VB1/9
FflX 1H

(Communications.
Paging Forecast Bright, with Widely Scattered Change..-

<RETUHN> =COHTIHUE LINE =FULL <P>REUIOUS <Q>UIT//EHPRESS

Number + II to MRRK for Order ( e x : HH) |

Strategic Fo]'cta5ting, L.1..C.

18/81/9

Mining the Open Source

141

years, permitting imperfect but useful searches of the paper

domain.
Indeed, one of the advantages of digitization is that it allows
you to locate material even if you are located in Podunk. We can
recall in our early days having to make trips to the Library of Congress in Washington, Harvard's Hollis Library in Cambridge, New
York's 42nd Street Library, and so on. Today, you don't have to go
to New York, Washington, Boston, or other large cities with archival
libraries in order to search their catalogsyou can do it from home
in Podunk. You never have to leave PodunkThis is real progress,
we suppose.
In the appendix is a list of libraries that we like to use. They are
all libraries accessible via the Telnet function on the Internet
(we provide addresses) and were selected for ease of use, size, usefulness, and so on. We have not included the finest library in the
country and possibly the worldthe Library of Congress
(locis.loc.gov)because its search engine is so bad that only the
federal government could have created it. Still, if you can figure out
how to use the search engineand you are better folks than we if
you cana world of riches will open up for you.

forking the System

W O R K I N G THE SYSTEM
Semiactive Intelligence Gathering

So far, we have focused on a particular class of material: open

source, public domain material that has been made deliberately
accessible by someone. What the Internet, on-line services,
CD-ROMS, libraries, courthouses, trade associations, and the rest
all have in common is that they provide information intentionally
to the general public. All of them have organized themselves, to a
greater or lesser extent, in such a way as to allow readers to find
what they need. In the end, the card catalog and Infoseek serve precisely the same purposeto allow you to search for information
using some formal system of organization.
This relatively organized realm does not by any means constitute all of the public domain. There is a vast amount of information
available that was not created with the intention of making it available to the public, but which can be legally and ethically accessed.
Some of this material exists in every information domain save, obviously, the innermost domain of proprietary secrets. Some of it is
electronic, some paper, some gossip, and some resides in the minds
of people. What all of this information has in common is that, while
it is available, it needs to be tracked down. Its existence must be
guessed at, inferred, logically deduced and so onbut it does not
exist with pointers available in any particular card catalog or web
search engine.
In fact, we can think of information in the public domain as
142

143

consisting of a series of networks. In the part we discussed in Chapter 5, the individual works are embedded in systemscatalogs or
databasesthat point to them and describe them. So, a book on
marketing bananas is listed in electronic catalogs under marketing
and bananas and any other appropriate descriptor.
There is also information that is not organized this way. Much
of this is located in networks of information that don't have external pointers, but which mostly point to each other. For example,
specialists in bananas know each otherthe banana world is a small
one. They also know of informationstatistics, studies, specialists
on various subjectsthat is available. They also personally know a
great deal about marketing bananas. There is, however, no database
or card catalog listing their names and addresses. They are well
known to each other, but are almost invisible to outsiders. You cannot access them passively. You actually have to go out and find
them, and that isn't always easy. Yet the payoff, if in nothing else
than printed information, can be spectacular.
To give you a sense of just how vast and even exotic this
domain can be, consider the following examples.
A professor, working under a grant from the National Science Foundation, has created a database displaying the consumption and disposal of car batteries. The database is in the
public domain, since it was created with federal money, but
you need to find out that it exists in the first place in order to
ask to see it. There are thousands of databases like this on
every possible subject in existence. More are coming into
existence every day.
The Qatar Chamber of" Commerce and Industry has produced a computer disk called Commercial and Industrial
Directory, which it distributed at the Middle East Economic
Conference held in Cairo on November 12-14, 1996. The
information is formatted as an Excel file. Over eleven thousand businesses are listed, along with their phone and fax
numbers, addresses, and type of business. You'll find everything from A Lady Tailor to the Subarah Shop for Selling
Old Equipment, all accessible in various ways.

144

THE

INTELLIGENCE

EDGE

The Compound Livestock Feed Manufacturers Association of

India has published a handbook that includes detailed statistics on Indian production and consumption of livestock feed,
dairy production, egg production, and so on.
Intersputnik, a Russian-based satellite consortium, publishes
a detailed guide to its products, including technical details
on its satellites and prices for services. You can lease hemispheric communications coverage from a Russian satellite at
$16 a minute. The cost goes down as the frequency of usage
increases. Of course, you might prefer a long-term lease.
Those numbers are here as well.
The Belgian Institut de Reescompte et de Garantie issues an
annual report detailing its own performance and its read of
the Belgian economyin both French and Walloon.
The Florida Department of Agriculture and Consumer Services annually publishes The Source, a complete directory of
Florida seafood suppliers, including what they supply. Three
companies supply cigar minnows.
A downsized engineer we know will provide you with detailed
information on selling thermostats to the U.S. Navy for far
less than any consulting firm you can hire.
A retired Israeli government official now living in the United
States will tell you more than you probably want to know
about doing business in Jordan in exchange for lunch. He's
bored and likes to talk.
All of this materialand vast amounts like it and probably
more usefulis still in the public domain. All of it is in digital or
paper format. All of it is obtainable. But none of it is being tracked
in a convenient way. That is the distinction that must be focused on
here. The geography of information conforms to the general geography we have been using; all of it exists in recognizable domains.
But none of them have been systematically "tagged" and entered
into general and accessible databases. These documents and the
millions like them are all riches beyond belief. All of them are
utterly legal uses of information. Indeed, in many cases, the pro-

Working the System

145

ducers and packagers of the information are eager to have you use
them and some may actually go to some effort to bring them to the
general public's attention.

THE PRIVATE PUBLIC DOMAIN

AND THE SEM1ACTIVE

There are many types of valuable business information that are not
to be found in even the best libraries and are rarely included in online services. Some kinds of nonarchived, noncataloged material
include:
Promotional material: Not brochures and come-onsalthough it
could include thatthis is a whole range of material whose production was motivated by parochial concerns. For example, let's say
that a developer wants to build a new shopping mall somewhere
and needs to convince the local authorities to permit the development. He creates a substantial study of the area's business environment, including demographic, environmental, and educational
data and other factors. In many cases, the study contains the best
data available on an area, and huge amounts of money have been
spent on it.
Obviously, the developer's motivation in producing the study
was promotionalto justify his own project. His conclusions will
reflect that. However, the data in the study and much of the analysis are both accurate and priceless; a superb window into a potential market. Now, the study has been distributed to town councils,
planning boards, and so on and is fully public, though it probably
has not found its way to the Library of Congress or onto a proprietary database. The developer wasn't motivated to do that and the
Library of Congress does not go hunting for this material. Nevertheless, if you can find out that it exists and find out who still has a
copy, it is probably yours for the asking. Promotional material can
include anything from proposals to planning boards to trade organization publications to brokerage reportsvast amounts of material produced by and about business but not readily available. For

146

THE INTELLIGENCE: EDGE

example, when large chain stores such as Wal-Mart or hotel chains

like Hilton decide they want to enter a market, they are frequently
required to produce studies for the local governing bodiestown
councils, planning boards, zoning commissionsshowing their
potential impact on the community. Apart from the self-serving
aspects of the report, the studies frequently gather some of the best
available information on all aspects of the local marketdemographics, traffic patterns, income distribution. All of it is in the public domain, all of it is free, and it would cost you tens of thousands
of dollars to come close to reproducing it.
Unpublished scientific material: Most scientific research carried out in
our universities is never published. Much of it isn't published
because it's worthless or someone has reached a conclusion that is
worthlesssuch as it can't be done. But much of it isn't published
because its findings weren't deemed significant enough or because
the professor was too lazy to whip it into publishable shape. Sometimes it will be published eventually, but currently it is unpublished
or has only been presented at conferences. There are real gems
here. Things not deemed significant by the scientific community
can mean a fortune to an entrepreneur. It is frequently well known
in the scholarly community that a given person is working on a project. People constantly pass around drafts of papers. In most fields
and among most people, there isn't an extremely high emphasis on
security or secrecy. Most scholars are delighted and flattered to find
someone taking an interest in their work. The trick is to find out
who is doing work that is of interest to you.
People: Which brings us to the most important potential source of
knowledge. People are the greatest repository of information,
much of it already transformed into knowledge. We have frequently
found that speaking to the right person for ten minutes is worth two
weeks of reading. If they don't know what you need to know, they
frequently know the person, article, book, or database that does.
Finding these people is much harder than getting them to talk to
you. This is particularly true of the greatest pool of untapped expertise around: government officials. These folks, contrary to popular

- Working the System

147

f mythology, know a tremendous amount, but they are buried in

a system that makes it hard to track them down. The same can
be said to a greater or lesser degree for banks, brokerages, and
hockey teams. All contain unexpected expertise alongside expected
expertise.
We are talking here about the majority of information and
human knowledge: the vast, unorganized soup that surrounds us
and is uslike The Force. Special effort needs to be taken to find
out about them, locate them, access them. But none of them can be
found passively. They won't come to you. You have to go out and
find them. This other material requires what we call a semiactive
effort.
Information and knowledge are frequently contained among
surprisingly small networks of people. Find one, you find
them all.
We have already discussed the difference between active and
passive search procedures. In the latter, it is not necessary to search
out information. The information comes to you via systems that are
already organized to be searchable. Therefore, it is not possible to
detect die search activity, something that may or may not be important to a business on any given project. In active searches, like the
submarine actively pinging its sonar to find other submarines, you
go out and find the information and it's extremely likely that you'll
be detected. Semiactive searching is an intermediate phase in
which the types of information you are looking for are quite similar
to those searched for in the passive mode, but in which purely passive techniques won't work. The semiactive search is not passive precisely because it is people dependent. The phone allows you to
conduct the search efficiently, but you are constantly dealing with
people, and as such, you are subject to detection. At the same time,
semiactive techniques don't require you to go to the extreme
expense of actually moving very far out of your officeif at all. That
holds cost and time down. This is critical for positioning on the
knowledge curve.
Obviously, anyone involved in a business Js already accessing

THE

148

INTELLIGENCE

EDGE

some of this material. The mere Fact that you are in the business
means you are on enough mailing lists so that organizations seek
you out. And the network you've already developed in your industrysuppliers, customers, bankers, friendsfrequently alerts you
to interesting information, or yields up the answers you're looking
for. But equally obviously, the informal methods you currently use
are neither exhaustive nor particularly useful when you go a little
farther afieldlike looking for potential customers in Qatar, suppliers in Florida, or financing in Belgium.
What we are suggesting is a systematic approach to an activity
with which you are fairly familiar. A systematic approach is critical
in order to assure you that you have accumulated the critical mass
of information needed to make a decision. The key to the system
really is to understand it as the opposite side of the natural information accumulation process we discussed in Chapter 4. There, ihe
normal activities of a business gather together large amounts of
wide-ranging information, much of it surprising and unpredictable.
Here the goal is to use the same informal networks as well as other
techniques to seek out, identify, and acquire extremely specific
information.
Specificity is the key to semiactive intelligence that should be
used only after you have a fairly clear idea of what you're looking
for. Sequentially, it must follow a careful review of information
already available in the company, as well as a systematic and aggressive accumulation of information from the organized public
domain. A careful study of this material should be made to determine whether you have already accumulated the information you
need to reach a decision. In many cases, you will find that the cost
and effort of gathering additional information may not be worthwhile.
Whan you have been to all the web sites, downloaded all the
articles, ordered ill the hooks and magazine articlesread
them! It will also help to think about them before you go off
looking for more information.
We have noticed an interesting tendency in our own shop: info
lusta disease that afflicts the industrious. When we begin a proj-

Working the System

149

ect, we start looking for information. Fair enough. We start gathering information. We look at catalogs, bibliographies, old studies,
our own databases. We start ordering tilings. We keep ordering
things. Things pile up, higher and higher. We order more things.
No one reads anything, theyjust keep finding more stuff. Then they
start calling around looking for more goodies. Finally, they run out
of time. They grab a few of the more likely looking pieces and try to
fake it. The poor souls who do this are suffering from info lust.
Notice, we are not calling it knowledge lustthere is no Faustian
legend here. They don't read the stuff, so they don't get any
smarter. Theyjust need information. We are gentle with these people, especially after the bills start coming in. Usually, warmth,
understanding, and counseling solve the problem. Executions are
not usually carried out until the second offense.
You must, absolutely must, read the material as it comes in, or
at least well before you start your second round of searches. You
must never pick up the phone to call around until after you have
absorbed the surge of passively obtained material. It is not only that
you won't know what you're looking for and might be wasting time,
but it is also directly harmful. Many times you find people who are
pleased to help you the first time you contact them, but they
become downright surly when repeatedly pestered. In many cases,
people are a wasting asset, at least in the short runanother call in
six months may not bother them as much as three calls in two days.
You really need to know what you're looking for before you start
making phone calls. Thus, a pause should take place at the end ol
the passive stagefor reading, thinking, figuring out whedier you
have everything you need, and deciding what it is you don't have.
Let's emphasize this point: The semiactive phase is missionoriented and task-driven. Its job is to fill in the blanks left over from
the passive phase. It is not the place to begin doing research nor is
it the place where the bulk of information is to be found. It is
| knowledge-driven. It should take place only after the information
gathered in the first phase has been transformed into knowledge.
^Otherwise, it is an opportunity for aimless schmoozing. But when
[it's knowledge-based, it's an extremely powerful phase of intelligence work.
We were once doing a study on the telecommunications indus-

TE L U I G E N C C

EDGE

try in the Middle East. Our client was another consulting firm that
was putting together an authoritative guide he hoped to the
region's telecommunications industry for his own clients. He
wanted to know how many satellite dish receivers there were in each
country at the present time. Simple question, not so simple to
answer. We hit all of our databases, but simply couldn't find the
answer. It didn't mean that there wasn't an answer, only that the
easy ways of finding it were not readily accessible.
We searched through our addled brains to find a way to answer
the question. OK, who did we know who knew about the Middle
East? Tons of folks. Who did we know who knew about satellite
dishes? Tons of folks. Who did we know who were members of both
groups? No one. This was turning into one of those killers that
cause our accountant to cry and gnash his teeth.
We had to think carefully. Who would we know in the telecommunications group who might be likely to know someone who knew
about the Middle East? OK, what group of telecommunications specialists was most likely to know people knowledgeable in the Middle
East? Bingo. Members of the U.S. military, people on the staff of
places like the Defense Information Systems Agency or the U.S.
Army's Communications and Electronic Command. Why? Because
the U.S. military spends tons of time in the Middle East. We called
a few and hit pay dirt on our third call. The guy over at the next
desk from the person we were speaking to had spent years handling
telecommunications for troops deploying in the Middle East. Did
he know the numbers? Well, the Defense Department had run a
study on telecommunications resources in the region. No, it wasn't
classified. Sure, he'd be happy to send me a copy. We didn't tell our
client that someone in DOD already had most of the study he was
doing sitting on his desk. It wouldn't have mattered, since his
clients didn't know about the DOD study and probably never
would. We wowed our client. He wowed his clients. Our accountant
smiled a rare event.

Working the System

PEOPLE:

FROM

151

POINTERS TO SOURCES

One way in which semiactive intelligence is used is by finding people who know of the existence of", and can locate, informally published documents. But in semiactive intelligence we begin to reach
the point at which people themselves start to become valuable.
When dealing with people, we've discovered, we need to increase
the care we take to be systematic and focused. This is tough enough
when dealing with books and articles, all of them chock full of not
particularly useful information. People are even harder. Not only
are they interesting in and of themselves, but they will talk about
what interests them rather than what interests youand you can
sometimes forget what it is that you called about.
At this point in the process, it is very important to return to
some of the basics we discussed in Chapter 3. The process of semiactive intelligence is, of course, unpredictable. Ever}' business, every
body of knowledge has a different geography of information, and
therefore requires very different actions. But the key in every case
is recognizing what is in hand, what remains to be found, and what
the most likely procedure is for finding it. It is absolutely essential
that the Mission Statement and the Information Inventory, now
updated to include information acquired during the passive phase,
be studied and that a new Information Requirements sheet be generated.

Is

As you move into the samiactive and active modes, you are
maving into a much more undifferentiated information soup.
Remain focused or you will find yourself lost faster than you
think. Do not stop to smell the roses.

I
I
I
$,'

We were once asked by a construction company to check out a

mid-sized food-processing company that had asked them to undertake an important construction process for a complex processing
plant. The project meant a substantial up-front investment. Quesn'on: What was the probability that the food processor would stretch
out the paymentsor worse, default on them? Legal protections
Weren't the issue. This was a dynamic construction company, which

152

THE

INTELLIGENCE

EDGE

meant that it was usually one step ahead of creditors. Cash flow was
everything and an interruption of cash flow would be catastrophic,
regardless of legal remedies two years down the road.
MISSION: DETERMINE THE LIKELIHOOD OF DEFAULT
OR STRETCHING OUT OF PAYMENTS

What did we know to begin with:

Our client was uneasy, having heard rumors that the
company liked to play games with its accounts payable.
Not being in the food-processing field ourselves, we
had only a vague idea of who the company was.
During the passive phase we found all sorts of information:
The food processing company was financing the plant
at interest rates of 16 to 18 percent.
The asset base of the financing was created through a
dizzying series of swaps between various companies
organized under the chairman and CEOlet's call
him Wildman Joeof the food-processing company.
The head man loved to sue and was regularly sued.
One of the reasons that interest rates were so high was
because everyone knew that, regardless of what assets
were pledged as collateral, default meant that everyone
would spend years in court litigating in the event of
failure.
The CEO of the company was being sued by his own
mother over another business deal.
It was clear to us that this company did not play well with others, and that the chairman had been out of school the day sharing
was being discussed. Nevertheless, it was a big contract and our
client did not want to pass on it if there was any way to ascertain the
company's likely course on this particular project. Would it be possible for us to figure out his intentions on this project, separate
from his general track record?

Working the System

153

Intelligence is a game of probabilities. We cannot look into the

hearts of men, and it was unlikely that Wildman Joe himself had any
idea what he was going to do with this project. He tended to go with
the flow. We didn't doubt dial he wanted to build the planthe
wasn't a straight-out rip-off artistbut we did know that if he ran
into trouble, the likely outcome would be to stiff everyone involved
and let his lawyers sort out the mess. The key, it seemed to us, would
be to try to figure out what the likely course of this particular project would be. This, in turn, depended on two different issues. First,
how likely was this project to succeed given die climate in the industry? Second, how healthy was this company internallycould it support this project?
Now, the food industry is, in general, a commodity industry.
Commodides tend to operate in a near-perfect marketplace. That
means that there are few secrets to be found. It is anyone's guess
what the weather would be like in Iowa next summer, or whedier
blight would get into the Russian crop. Indeed, commodities were
so perfect a market that a contrarian position is probably the best.
Unlike research in a highly technical industry, like computers or
Pharmaceuticals, where the opinions of the best scientists and analysts are a pretty good gauge to what is likely to happen, specialist
opinions in a commodity area are extremely ambiguous indicators.
Using semiactive intelligence in this casecontacting key commodities brokers and specialistswould yield interesting information perhaps, but in the end we would not be able to rely on the
consensus in determining our own recommendation. If anything,
we would tend to go against the consensus. Even the best track
record is no guarantee of future resultsas the prospectuses tirelessly and accurately tell us. So calling the guy with the best track
record won't give us very much that we could rely on. Frankly, if the
food market collapsed unexpectedly, all bets would be off. Our
client was prepared to live with that risk. What he needed to know
was whether Wildman Joe could, under the best of circumstances,
pull the project off and whether he had something weird up his
sleeve.
We clearly needed to focus in on the company. Some questions
emerged. We needed to develop a general read on the internal

154

THE

INTELLIGENCE

EDGE

finances of the company. We needed to figure out what part this

particular project played in Wildman Joe's overall plans. We had
found what there was to be found in the public domain. It was
murky. How could we find some clarity?
In analyzing the project, we focused on the most salient point:
Wildman Joe had borrowed a heck of a lot of money at an
extremely high interest rate to finance the project. Two questions:
(1) Why had he gone that far out on the yield curve? and (2) What
would be the general result on his businesses if he did not bring this
project home? Now, we could speculate and we could guess, but at
the end of the day we had to have the answers to these questions.
Indeed, the two questions were really one: What did this project
its success or failuremean for Wildman Joe's personal fortunes?
So, was this another of the schemes for which he had created
thirteen escape hatches in the event that anything went wrong, or
was this something that was of substantial importance to him. That
would tell us whether he was going to jerk around the contractors
because the one thing we knew beyond any shadow of a doubt was
that Wildman Joe took care of Wildman Joe.
In the end, it was all as simple as this. But who would know the
answer? Or more precisely, who would know the answer and would
also be willing to tell it to us? Yes, we were tempted to call his
motherbut after all, a mother who would sue her own son is not
necessarily going to be straight with us. This was a life-or-death issue
for our client. So where should we turn?
What did we know? First, we knew that Wildman Joe spent bis
life in court and that he had lots of enemies. Bingo! In the United
States, there is a lovely thing called discovery, where people suing
each other get to paw through each other's dirty laundry. Some of
this finds its way into the public domain, but a lot of itparticularly
the impressions drawn from the materialwind up embedded in
the heads of litigants and their lawyers. And from our passive work
on Lexis, we had a list of litigants and their lawyers a mile long. We
were getting somewhere.
Now, lawyers tend to be a closed-mouth lot when they're talking about their clients, but they are not always so circumspect when
they are talking about their adversaries. And the litigants them-

Working the System

155

selvesthe poor suckers who had been worked over by Wildman

Joewere not only not ethically bound but usually in a mild rage,
to say the least. Of course, the list of possible sources was pretty
long, so we needed to define what we were looking for:
1. We were looking for people whose litigation would have been
directed at Wildman Joe himself, rather than at one of his
companies. This was designed to ensure that we would be
dealing with people who had glimpsed his personal situation
during discovery.
2. We were looking for small businesses rather than multinationals for two reasons. First, in our experience, smaller business people and their attorneys tend to be more garrulous.
There is simply a lot more emotion built into a case in which
your whole existence is at stake compared to a case whose
outcome won't even be noticed on the bottom line. Second,
they were more likely to have retained the nuances of the
matter in their minds and to have tried to go for the jugular.
They would probably know more.
3. We wanted more recent lawsuits, but not so recent that discovery had not yet taken place.
OK, we had a list of cases from LEXIS and Westlaw, the names
of clients and their lawyers. It was time to pick a few. We found
about twenty and let our fingers do the walking. We were very frank
about why we were callingand very cagey about who we were calling for. Our client did not wan,t Wildman Joe to hear about what he
was doingcertainly reasonable. Once we explained what we were
|;,, doing our only task was to relax and take it all in. The litigants
| themselves immediately started pouring forth vitriol. The lawyers,
having a reputation to live up to, remained totally professional for
|, about ten minutes, and then their own anger started to rise.
Some important things to remember when manning the tele| phones, probably pretty obvious but worth stating anyway:
The person to approach is the one with an interest in talking
to you. Remember, that interest is not defined as just

156

THE

INTELLIGENCE

EDGE

financialpsychological interest is at least as important

as financial interesteverything from ego to sympathy
comes out.
Define the likely mind-set of the listener and play to it. Here,
a potential victim (your client) is looking for help in not
being burned and, in turn, you are offering them another
forum to tell their tale of woeand maybe do a little harm
to an enemy.
Select the appropriate person and/or persona for the phone
call. Don't send a young kid to call a senior partner at a
law firm.
Allow the speaker to unload on you. Act as a bucket to catch
what conies out and sort it later if necessary. But always keep
in mind what you are after.
Allow him to identify with you. You must not identify with
him. You are playing your game. He is playing his. Remember: Our reason for calling is to find a way to do business
with Wildman joe, not merely to celebrate his wickedness.
In doing semiactive intelligence, the main cost is time. You
have to debrief your source at his speed. Keep in mind that
his motivation is not to solve your problem but to satisfy a
peculiar psychological itch of his own. Patience is the price
you will have to pay.
But it was Wildman Joe's meanness that was our pay-off. He
had screwed so many people, so royally, that all we had to do was listen and wait. In the end, the details of his personal empire were
clear to us. Wildman Joe was in fact in personal trouble. If this project failed, he was personally finished. When we laid it all side by
side, what we saw was a personal life and personal finances in massive disarray. We were not told this by any one person. Rather, each
of them, in the course of telling us how they had been burned and
what they found out during discovery, painted for us a collective
picture of real financial trouble for Wildman Joe.
The fact was that, unlike most cases, he was personally in more
trouble than his companies were. He had structured the deal in

Working the System

157

which our client was involved so that he could personally revive his
finances. He needed a home run and this project was it. Now, in
most cases, this would be a recommendation to avoid involvement,
and had our client been an investor this might well have been our
recommendation. But he was only a contractor looking at a potentially lucrative contract. The paradox was that, in this case, Wildman
Joe's personal financial problems made this a more secure play. He
was using his company's resources to rebuild his personal finances.
He needed this project to work. He was far less likely to play games
with it during the early phases than jn his other projects.
Our recommendation was to go forward with the deal, trading
some long-term benefits for short payment schedules to protect
against the project getting into financial trouble, but expect Wildman Joe to work hard to get the project done. Moreover, he needs
you simply because a lot of other folks are not prepared to do business with him anymore and you are. Our client took our advice, did
the deal, and got what was owed him. Better still, Wildman Joe, a
born-again nice guy after his brush with disaster, throws as much
business to our client as he can handle, since he stuck with him
when no one else would. And he could stick because he had the
information needed to make an informed, rational decision.
This was a case in which information that we obtained passivelycourt recordsallowed us to determine clearly what we
needed to know, and also allowed us to locate individuals who had
the information and knowledge we needed to make a recommendation to our client and for him to make a decision. In many, if not
most, cases, this is true. Formally published, informally published,
and personal knowledge are so intertwined that accessing one part
of the information network will allow you access to all other subnetworkspeople, unpublished papers, and so on.
There are, of course, places where this doesn't work, particularly in areas that are new to you or in which you have allowed your
expertise to atrophy over time. We once had an assignment to produce an ethnic map of Slovakia. This was an intelligence project in
the full sense of the term, requiring us to pull together disparate
information and merge diem together with technology. Our problem: Since this took place right after the collapse of the Soviet

158

THE

INTELLIGENCE

EDGE

empire in Eastern Europe, we didn't know where to start. How

could we get the information on ethnicity in a country where all
data was a tightly held state secret and in which ethnicity was an
explosive subjectand on which very little had been written in the
West because of lack of interest and lack of access. Our passive
searches hit a blank wall. We didn't even have a Slovakian telephone book, let alone a Rolodex full of numbers.
A new technology proved extremely useful, providing us with
a new form of semiactive intelligencethe Internet. As discussed in
Chapter 5, this part of the Internet is not the graphically spiffy part
that most of us are getting used to, but the newsgroups that have
proliferated. Under Slovakia, there were several newsgroups,
including one called Alt.Soc.Slovakia. We looked at that and found
a rich discussion on Slovakian society. So we posted a question:
Does anyone out there know of anyone doing work on mapping
ethnic groups in Slovakia? We also posted this on the Hungarian
newsgroups, since ethnicity in Slovakia is very much a Hungarian
question as well. This not only started off a new thread on both
newsgroups, and prodiiced for us manifestos on the injustice of one
side or the other, but tapped into a network of Hungarian and Slovakian scholars working on the issueas well as some government
officials wanting to know why we were interested. They turned out
to be quite helpful in the end.
Notice the geographical shift involved here. Passive intelligence concentrates on electronic and paper. Semiactive still works
these domains, looking for unindexed information. But it also
moves in a much murkier domain: gossip and to some extent the
gray area between gossip and proprietary information. It is vital to
bear in mind that the realm of gossip in information soup is soup
information floating intermixed with other information and even
disinformation. The gray area is just that, gray on gray on a foggy
day. It is a realm of unclarity and uncertainty, where the motives of
all the players are unclear and in which the information you receive
is difficult to verify.
The increasing uncertainty inherent in dealing with these interior realms demands increased disciplinea point made over and
over. Unlike the internal accumulation system, which is intended to

Working the System

159

capture and sift lots of gossip, semiactive forays into gossip collection and into the gray area requires a high degree of clarity. In our
experience, premature entry into this realm can be disastrous.
Without clear grounding in the general area and in the questions
you need answered, you will inevitably run into increased rather
than decreased confusion. At least as important, you can quickly
become a victim of disinformationbeing given answers that, were
you more sophisticated, you would quickly recognize as, in some
sense, false. Without preparation, trading in gossip or semilegitimate data is a nifty way to become a victim.
A step-by-step approach is recommended:
Identify the missing pieces: Study the board. Assess what you have and
what question remains to be answered. Determine how valuable,
how necessary that information is in giving a final recommendation
or making a decision. Decide, beforehand, how much time and
money you would be willing to spend on it.
Identify the, likely location of the information: In one of our examples,
the likeliest location for what we needed was in the heads of lawyers
and clients, to be contacted by phone. In our other example, it was
in the heads of scholars, to be initially located via the Internet. By
the time you have reached the semiactive point in your research,
you should know where to access the needed information. If you
can't, you have more homework to do. This is a good test as to
whether you're ready for transition.
Task-suitabU personnel: You may not want to use your most intelligent
people here; they may, after all, be social asses. So focus on the most
congenial people you have. Indeed, if you are going to do this a lot,
you will want to make sure you have a group of diverse personality
types for each missionor at least people with the ability to act. In
some cases you want mature adults, male or female, to do the querying. In other cases, you want the young and innocent. It varies by
case and context. Remember, you are dealing in gossipa very personal and even intimate social process. If you need information
from the jefe of a Guatemalan village and you figure that he

160

THE

INTELLIGENCE

EDGE

responds best to an implicit sense of danger, don't have a nineteenyear-old woman contact him. On the other hand, if you want information from someone who will respond better if he feels he's doing
someone needy a favor, don't send an ex-cop to ask questions. Gossip is social in the sense that it is designed to elicit information
using socially powerful cues. These cues vary in time and place, and
cannot be subject to political correctness. We are in the real world,
which is inhabited by some very real and powerful stereotypes.
There are still things like pretty girls and tough guys in this realm.
That is the way it isand our apologies to political correctness.
Define means of contact: Today, there is a wide range of mechanisms
for contacting people. We have gone from mail to telephone to fax
to e-mail and to all of the more exotic emerging means. If you are
scoping out things in Silicon Valley, the web is both desirable and
logical. But if you want to contact owners of small clothing stores in
midwestern towns with populations under 25,000, you will probably
want to choose more traditional means. Always bear in mind that as
you move deeper and deeper into the information structure, personal contact becomes more and more important.

CONCLUSION: U N R A V E L I N G THE N E T W O R K

It can begin simply: checking the bibliography of published works

to locate publications you haven't found yet or which are not available in libraries. Yon find out the telephone number of the person
who wrote the publication and call him. He turns out to have several other unpublished manuscripts and will send them to you. He
also knows two or three other people you really ought to talk to; feel
free to mention his name. By the time you're done, you have
absorbed an extraordinary amount of information and knowledge
for shockingly little money or effort.
We are quite serious in speaking of the compact vastness of
networksand we're not just trying to be cute. Expertise is rare.
One of the characteristics of the expert is his comprehensiveness:
He knows what he knows arid he knows other people who know

Working the System

161

things as well. His knowledge is vast in relation to the subject, but it

is compact in terms of time and utilization. He makes his living by
being efficient in utilizing his expertise. In tapping into his expertise, you find vastness and comprehensiveness, but also compact
efficiency.
Experts leave tracks; they are never invisible. We made this
point when discussing the inside of the company and must repeat it
here. They publish papers, give interviews, serve on boards. Now,
one problem you have is that the most prominent expert is frequendy the least likely to have the time or inclination to help you.
But the most prominent is not necessarily the most knowledgeable.
Less prominent, younger, hungrier specialists not only know as
much but are frequently more motivated to help people in order to
build their own unique network and expand their own horizons
plus build their egos. Our strategy is usually to track down a member of the second tier in the profession. He is more likely to be
cooperative and quite probably more knowledgeable.
It is the first step that is hardest. Assume that you are entering
a field you know nothing about. How should you proceed? Well,
assuming that you've finished your passive searchand we assume
that because you have absolutely no business doing semiactive
research until you haveyou must now define what you still need
to know. Having done that, go back to the publications and, from
what you've read, extract the names of experts in those areas. That
will do the trick 90 percent of the time. Certainly, some people
won't talk to you, but we have found this to be a surprisingly small
percentage. Sometimes you may have to pay for their time to even
talk to themfrequently trtie with lawyers and consultants. Fair
enough. Calculate the potential value against the cost and make
your decision.
There are cases where publications won't work. Here, unraveling the network becomes more difficult, but not impossible. We
favor three approaches at this point:
Locate a likely unit of government. For example, if you are in
the basketball industry, and you are looking for new markets,
a logical place to begin is your nation's Education Depart-

162

THE

INTELLIGENCE

EDGE

ment or Ministryspecializing in physical education. If you

don't know who to ask for, explain what you want to the
receptionist. They will want to get rid of you quickly, and the
best way to do that is to pass you on to someone else. Eventually you will find the one person in the building who worries
about basketballs. He will tell you about all sorts of studies
and meetings and people. The very bloat of government can
work to your advantage. And most government officialsparticularly mid-level officials not used to putting up with public
abuse on a daily basisare actually reasonable, affable
people.
Locate a semipublic organization, such as a chamber of commerce, trade association, trade union, someone not associated with a particular business but with the industry in
general. These are superb places for starting to unravel the
network.
Universities are useful for more abstract or abstruse areas
of knowledge, rather than concrete areas of marketing
or finance. But if you have a question on penetrating the
Peruvian jelly bean market, you may not find a jelly bean
expert at the university; however, you will probably find a
Peruvian expertand that may begin to unravel the
network.
One additional point here: The contacts and knowledge accumulated in these forays are remarkably and unexpectedly valuable.
No phone call should ever be made without inputting basic facts
onto a contact sheetpaper or computerized. It is gratifying when
a failed contact turns out to be of extraordinary value.
Never throw anything sway. The effort of saving ene hundred
useless contact records is paid for by the use ef one
unexpected contact.
Locating and unraveling the network is usually a fairly easy
processsurprisingly easy given the amount of value embedded in

Working the System

163

these networks. Rarely, but sometimes, all of die passive and semiactive intelligence won't do the trick. This is usually a time to hang
it up. Sometimes, however, you can't let go. This brings us to the
rarest, but for some perverse reason sexiest, aspect of intelligence:
the active operation.

When You Absolutely, Positively Have to Know

WHEN YOU ABSOLUTELY,

POSITIVELY HAVE TO KNOW
Active Intelligence Gathering

There are times when you desperately need information, times

when the normal constraints of cost, time, and even risk don't
apply. Times when the normal process of information collection
and knowledge production simply can't tell you what you need to
know. These are the times to use active intelligence collection. Let's
say, right from the beginning, that, risk aside, the efforts and costs
associated with active intelligence gathering are such that most
businesses, in the course of their entire history, will never need to
resort to such means. In a sense, what most people see as the heart
of intelligenceactive meanswe see as not only peripheral
but rare, usually costing more than it's worth. But rare does not
mean never.
Active intelligence involves more than getting out of your
chair and out of the office, although it probably means that as well.
Active intelligence means going out after information you need
once you've determined that it is simply not in the readily accessible domainsor that it is difficult to track down using the tools discussed under passive and semiactive intelligence.
This information is not so much illegal to gather as it is
shrouded in secrecy. This is an activity where the borders are more
than a little hazy. Put it this way: Information can be owned by
someone, and taking it constitutes theft. It is, however, possible to
legally pierce the veil of secrecy by observing perfectly legal actions
164

165

and inferring from them underlying reasons using other information you have obtainedalso legally. If this all sounds a bit cloudy
and you are unclear as to what we are saying, then you are on the
right track, because this area of intelligence is murky and it's never
easy to follow what's going on.
Take some pretty famous examples. During the Cuban Missile
Crisis, reporters noticed limousines coming and going from the
White House. As old hands, they knew this was unusual activity.
From the license plates, they could figure out who was coming and
going. By noticing that the plates belonged to both Defense and
State Department officials, it was clear that a foreign policy crisis
was brewing. It didn't take much to figure out that it concerned
Cuba. The same thing happened during the Gulf War. Reuters
reporter Jacqueline Frank actually wrote a piece entitled "How
Tense Are U.S. War Planners? Just Count the Pizzas." On January
16, 1991, the day before the air campaign began, she noted that on
the previous night, Domino's Pizza had delivered fifty-five pizzas to
the White House between ten P.M. and two A.M. The average White
House order during that time frame was five pizzas. At the Pentagon, just over a hundred pizzas were ordered; at State, seventy-five,
up from one per night. The CIA, using their massive resources to
locate the leak, stopped ordering from Domino's. Frank shrewdly
grasped that the massive increase of pizzas being delivered late at
night meant that the number of people working late had increased.
She also understood that there was probably a reason lor this and
that this reason was probably not associated with a study of the
National Park system.
Now, the fact that there was a Cuban Missile Crisis was a secret,
as was the timing of the beginning of the air war on Iraq. Anyone
tapping phone conversations, reading secret documents, or blackmailing participants would be committing a serious crime. What
was not a crime, in the United States, was observing an uncontrolled emission of informationthe physical, visible presence of
official limos or of Domino's Pizza delivery cars. Those emissions,
placed into context by individuals possessing highly nuanced information from years of covering the White House, generated knowledge of what was going on that was every bit as clear as if the

166

TH

NTE

GE NCE

EDGE

president's press secretary had made an official statement. It was

possible to penetrate the innermost secrets of the White House perfecdy and totally legally. But you literally had to be there to see the
limos and delivery trucksor at least contact Domino's directly and
hope they answered your question. More important, you had to
have a massive base of information on which to draw to understand
the meaning of the limos.
Presence is the key to active intelligence, and what makes it so
expensive. Stationing a reporter at the White House is expensive,
but it's a no-brainer for a large news organization, a decision that
pays for itself in many ways other than the occasional intelligence
coup. The White House is an obvious location for a news organization to place a reporter. The problem is much greater for a business
since it is not in the intelligence-gathering business and can't cost
out the price of an agent over other, profit-making, projects, like
the evening news.
But presence is often indispensable. Sometimes the reason for
this is to pick up information before it diffuses into the public
domain. Other times, this is because the information will not diffuse and will only be available to those present. And then there are
the times presence is needed because information has to be coaxed
outin effect, an emission has to be generated. Whatever the reason, active intelligence requires presence by someone who is able to
recognize the difference between the significant and insignificant.
That's what forces the price up. Expertise doesn't come cheap,
especially when it has to go out in the field.
IB active intelligence, if you have to ask the price, you
probably have mare sense than to try it.
It is sometimes possible to find these emissions passively. However, in matters of vital interest, emissions are guarded extremely
carefullyparticularly those that are likely to move into the electronic domain. Often such emissions show up in the realm of gossip. During mergers and acquisitions, for example, gossip is
frequently generated prior to official announcements due to accidental or, at times, deliberate leaks. And there are times when pas-

When You Absolutely, Positively Have to Know

167

sive intelligence will pick up a nugget that is overlooked by others

but from which an experienced hand can draw inferences. But
sometimes, the intelligence coup does not derive from a new bit of
information available only to you but from the ability to analyze and
understand a generally available but neglected bit of information.
At best, active intelligence gets hold of useful bits of information sooner than would otherwise be the case. This is the core issue
of active intelligence: positioning yourself on the knowledge curve.
Active intelligence can provide you with two advantages. One is simply accessing information that would otherwise be unattainable.
The other is accessing information before others have it. The times
when the potential payoffs for this kind of information are worth
the cost are extremely rare.
The rareness of rational active intelligence in business is made
rarer still because active intelligence has the highest probability of
failure. As the White House examples show, a good deal depended
on lucknoticing the events and understanding them. In many
cases, active-intelligence projects are undertaken without a clear
goal, which substantially increases the need for luck. After all, sending someone out to find out somethinganythingabout the
competition merely increases the opportunities for trivia while dissipating the energies of the collectors.
Nothing is more important in active intelligence than defining
precisely what it is you need to know, how you expect to find it out,
and what the financial value to your business will be. Virtually anything can be found out with enough time and enough money.
Merely beginning the project starts to run the clock at an exorbitant
rate. Our rule is that 90 percent of all information you need can be
found through passive and semiactive intelligence. Of the remaining 10 percent, you will find that 90 percent of the desired information is simply not worth it. By the time you get it, it will be too
late. By the time you pay for it, it will have destroyed your profitand-loss sheet rather than having enhanced it. That means that
| active intelligence will be justified in no more than 1 percent of all
|, casesand frankly, we diink that is a very generous assumption.

168

THE

INTELLIGENCE

EDGE

THE A C T I V E INTELLIGENCE OPERATION

Assume that you have encountered a business problem that

requires information not accessible by passive or semiactive means,
and assume further that you have determined that securing this
information is of enormous value and you are prepared and able to
put substantial resources into the project. Now what?
After you have precisely defined what it is you need to know,
the next step is to try to figure out why you haven't been able to find
it in the first place. Many times you discover that the reason you
haven't found it is because it doesn't exist. At other times you discover that it exists but you just haven't looked in the right place.
Sometimes, when you think about it, what you really want is not so
much information but analysis. This is a particularly nasty mistake
to fall into.
For example, let's say that you want to know if another company is going to produce a product that would compete with one of
your own. You know the company is capable of doing it, and you
think it would probably benefit them. Many decisions you are about
to make depend on their plans. It is something you really need to
know. The problem is the information may not exist because the
decision makers at the other company have not yet made a decision. That doesn't mean they won't decide to go ahead, so you can't
discount it in your planning. But they may decide not to produce
the product, so you can't discount that either. You need to generate
a Plan A and a Plan B to cover both alternatives. And often you can't
wait for a definitive answer to actby the time the other company's
CEO makes his announcement, it will be too late to respond. You
need an answer as early as possible.
An active operation that comes up with nothing may not be a
good sign. You are now in a classic dilemma of intelligence, one
that is particularly painful given the high costs of active intelligence.
A negative answer may mean that he is not doing it, or it may mean
that your intelligence operation has failed. A positive answer may be
a false positive, or intentional or accidental disinformation. This is
the classic problem of active intelligence. Any answer you receive
may or may not be true. And absolute verification is always difficult.
In the end, you are forced to go with probabilities in matters where

When You Absolutely, Positively Have to Know

169

nothing less than certainty will do. In increasing your information,

there is no certainty that active intelligence will increase your
knowledge.
The goal, of course, is to minimize this chaos. Before you
begin the operation, you must determine what sort of information
will satisfy your needs. In this case, what is really needed is analysis.
What is the best move for your quarry to make? In the past, was he
someone who made the best move or did he tend to screw up? What
are the factors that shape his decision-making process? None of
these are part of the intelligence-collection process. All of these use
intelligence products to carry out analysis, and in the end, all intelligence culminates in analysis. In both government and business
there is an ongoing fear of pitting your mind against your competitor's mind. It is much easier to look for more information. The fact
is that you may have in hand all the information you're going to
need. There is a tendency to push gathering much too far.
However, if it does turn out that additional information is
required, then you have to determine several things:
Where is the information located?
What is the form of the information? Is it a document? A
conversation? An idea in someone's head?
Is the information you need directly accessible?
If not, what emissions might permit you to infer the information you need?
What resources must you allocate and how should they be
deployed to capture the information or the emission?
There are two kinds of resources that can be devoted to activeintelligence projects: technical collection systems and human intelligence systems. In the case of active intelligence collection, human
intelligence becomes much more important. Most technical collection systems that are used by national intelligence agencies are illegal for use by private citizens. Signal interception, for example,
which is done regularly by the U.S. National Security Agency as well
as by other intelligence agencies, is not something most companies
can legally undertake, outside their own premises at least.

170

THE

INTELLIGENCE

EDGE

Check out the rules for the specific country in which you are
working. Although there are numerous parts of the world where
photographing transportation or industrial facilities is illegal, there
are just as many countries where whatever can be seen in public can
be recorded. Photographic intelligence, while often not very useful,
can be extremely rewarding in specific cases. In countries where
photography is not a problem, it can open the door to any number
of possibilities. Take the following examples:
A critical issue for business and government alike is the trend
in the price of oil. During the early 1980s, a friend was flying
us to Long Island, east of New York City, in his private plane.
While flying from New Jersey to Long Island, we crossed over
the Atlantic approaches to New York harbor. I was startled to
see numerous oil tankers, none showing wakes, clearly
anchored for scores of miles outside New York harbor. I
researched the question and discovered that oil companies
frequently ordered tankers to stand at anchor waiting for
shifts in the spot market. Looking at that line-up outside the
harbor, we realized that the sellers were not entirely happy
with the price they were getting and were hoping to see it
rise. Also, so many tankers were piling up that it was unlikely
that the price would rise. In fact, we were seeing the beginning of the 1980s oil glut- Just looking at what was happening
in New York harbor could not justify my going short on oil
futuresafter all, a temporary breakdown in oft-loading facilities could have been the cause. Today, however, it is possible
to monitor off-loading activity in all of the world's leading
portsYokohama, Rotterdam, Houston, and so on. The end
of the Cold War has made commercial access to satellite
imagery available at surprisingly low prices. For example,
Russian Spin-2 satellite imagery with two-meter resolution
you can see carscosts $300 for a 2km x 2km fragment, supplied on CD-ROM or 8mm tape. The French sell synthetic
aperture radar shots at fine resolution for $1,850 for a 100km
x 100km swathjust right for tracking shipping. A rush
order can be processed in under a week, but costs extra.

When You Absolutely. Positively Have to Know

171

Satellite imagery allows you to monitor large-scale develop

ment projects that are inaccessible by other means. Suppose
a competitor is building a major new manufacturing facility
in China, in direct competition with your company. It is obviously important for you to know how construction on that
project is going. A satellite image of the site, taken in various
spectra including infrared and visual light, could very well
tell you where the project standsand whether you will have
to worry about new products in the next month or next
decade.
An old trick, and one not requiring satellites, is to take
regular pictures of the parking lots of competitors. A
tremendous amount can be learned by comparing pictures
of parking lots over time. Obviously, you can get a real
sense of the workforce, and whether the facility is operating
at the same level as last year can easily be seen by the number
of cars parked at various times of the day. At smaller companies, you can tell something about the relative prosperity of
the company: Are there a lot of Porsches in the parking lot
or old Chevrolets? This is active, technical intelligence in
the best and simplest sense. But just looking isn't enough.
Photographs are needed to give you historical context and
benchmarks.
The visual is the most legally accessible spectrum in active
intelligence and one of the reasons that, for business, much active
intelligence is done by humans. Human intelligence, or HUMINT,
is generally carried out with the eyes and ears, picking up information that is passed into the public domain and is therefore accessible. Following someone, overhearing a conversation at the next
table, staking out the entrance to a building are not illegal. They
may not be particularly ethical, but you will probably not be taken
away in handcuffs if it's done with common sense.
Technical collection systems are very sexyrarely useful, but
irresistible. Come to think of it, that's what sexy means.

172

THE

INTELLIGENCE

EDGE

Human intelligence, on the other hand, is more simple than

sexy. We once had a case in which we needed to check out a Russian company. The company, whose headquarters was in Moscow,
had approached an American firm with a joint venture proposal
that was quite intriguing. Obviously, the American company wanted
to know more about the Russian company. Passive intelligence, with
any degree of detail, on Russian companies is quite difficult. They
are about a generation behind us in the information-production
industrymeaning that they do not yet use computers for typesettingwhich means that most of their reporting is inaccessible electronically. What little we could gather passively on Russia came
through Western reporters writing on Russian business, and this
tended to focus on the activities of very large Western companies
and their Russian allies, not on independent Russian firms.
Phone callssemiactive collectionwould not help either.
We needed to know simple things, like the kind of office this company had, who works there, who is their banker, and so on. We
needed to see the facility; we needed to talk to people who were
unlikely to talk to a stranger over the telephone. In short, we
needed someone there to do some legwork. It really didn't take a
great deal of skill or even reliability; it just needed a body with a
modicum of intelligence. Sending someone to Moscow was
extremely expensive and no one on our staff, at that moment, could
pass as Russian. This was quite important, for a native gets a different response than a foreigner. We were looking for texture here
how solid was the business? A moderately intelligent and reliable
Russian would probably serve us better, and certainly cheaper, than
one of our own employees going over there.
This poses the key problem of human intelligence: finding the
right person. Sometimes this is an impossible task; other times, such
as this one, it appears to be impossible but is really not so bard.
Finding what we might call a generic human, Muscovite flavor,
while sitting in an office thousands of miles away would daunt many
people. Most large companies turn to professional investigative
agencies, which charge an arm and a legand usually a lung as
well. That level of expertise was neither required nor justified for
this project.

When You Absolutely, Positively Have to Know

173

The point of the play Six Degrees of Separation, about a young

man claiming to be the son of Sidney Poitier, is simple: No one is
ever more than six links away from any person. Everyone knows
someone who knows someone who knows someone...until, within
six links, you can get to the person you need. Now, as an empirical
matter, we don't know if this is trueit certainly sounds goodbut
it does point to the fact that there are an extraordinary number of
linkages at our disposal.
With this in mind, we hit our friends and acquaintances. We
asked our staff if they knew anyone who lives in Moscow. Within
twenty-four hours we found two U.S. graduate students living in
Moscow, a biologist working there, and the in-laws of one of our secretaries who were Russian nationals living in Moscow. Our secretary's in-laws were retirees from one of the ministries who spoke
some English. We were on the phone with them within an hour.
They were absolutely delighted to do some work, earn what
was for them a fabulous amount of Western currencyand for us,
frankly, not a whole lotand help their daughter-in-law as well.
They knew the address wellit was a short distance from their
home. They went to it, reported that the "headqttarters" was in a
shabby apartment house; the telephone and fax numbers were private rather than business exchanges; and no such company was
known to the building superintendent. They also noted that there
were no foreign private license plates on cars parked around the
areaa sure sign that this wasn't an area where internationally connected businesses were established. With a high degree of confi; dence we were able to report to our client that this company was,
'T like many struggling start-ups in Russia, not more than a name and
, a P.O. box, and that unless they were willing to get in there, put a
dot of money and expertise into the individuals involvedwho
[seemed pretty sharp, we were toldthere was not much at
fthe other end to count on. Bad news, but better to hear it early
than late.
This was a case of active human intelligence. It was not rocket
[science. It was not a case that involved great cost or risk. It was simIply the only way to handle a case where there was little passive inteljjence to hold on to. The hardest part here was finding someone

174

THE

INTELLIGENCE

EDGE

to do it. However, even small businesses can readily find someone

to check something out so long as it isn't extremely complex,
extremely sensitive, or in some truly godforsaken corner of
the world.
Assuming that your personnel network doesn't turn up someone to do your legwork, here are some suggestions:
If you can't find someone to check something out in your
own country, you had better improve your social life fast.
Universities are tremendous repositories of people who know
people. This is particularly true of the language and social
science department. If you want someone in Paraguay, call
the secretary at the Spanish department, history, anthropology, or political science department. Ask for their Paraguay
specialist. After they stop laughing, ask for their Latin Americanist, any of them. Ask them if they know anyone specializing in Paraguay. They'll know someone. Get a number and
an introduction. The Paraguayan specialist will know a dozen
people happy lo earn some money. He may be going there
next month.
Call the U.S. Commerce Department and ask for the person
specializing in Paraguayan trade. Call the U.S. State Department and ask for the person on the Paraguayan Desk. In the
United States and most Western countries, someone in business can readily access even the most senior officials working
a country. You will be surprised who you can get to speak
to if you exercise authority, courtesy, and patience. But
usually, with many requests, lower-level officials will be more
useful.
Call the U.S.-Paraguayan friendship association, chamber of
commerce, or whatever. You can probably find them on the
web or by calling the nation's embassy. Don't rely on the
embassy itself for information on particular companies native
to that countrythe con artist might turn out to be the
ambassador's brother. But you can get quickly plugged into a
suitable relationship.

When You Absolutely, Positively Have to Know

175

Sometimes there are cases that are more sensitive. A client

company we were involved with was interested in making acquisitions. One company they were looking at, in New Hampshire, was
an old family firm, where the head of the family had recently passed
away. A son-in-law was now running the company. Our client badly
needed to know the son-in-law's intentions. Was he planning to continue running the company; was he thinking of going public; was he
open, eager, or even desperate for a buy-out? We knew that the sonin-law had recently been a senior official at a major computing
firmnot at all this company's lineand that he had left his job to
handle the family business. He may have been delighted to do so,
having waited for decades for the old man to die. Or he might have
done it rcluctandy, forced into it because there was no one else in
the family to take over, and eager to get back to his old job. The
facts we gleaned using passive and semiactive means leant themselves to either interpretation. We needed to know for sure.
There were two approaches. The first would be to look at his
finances and those of the company. Both were extremely closely
held. It was simply not legally possible to see them. The other
approach was to get a sense of the guy's thinking. We didn't know
him, and we didn't know anyone who knew himcertainly not well
enough to tell us something this important. But in one small article
we picked up, we found that he was a member of a country club,
and a fairly active member at diaton the board and so forth. He
would be well known at that country club and organizations like
that are hotbeds of gossip.
The question we now had was: Did we know anyone who might
be active in diat particular country club? Given the focus of our
attention, our needs were pretty well defined: We needed someone,
fairly well-to-do, who lived in or around that New Hampshire town.
The odds were excellent that if he didn't know the guy, he would
know someone who did, and if he didn't belong to the club, a
good friend of his probably did. So we needed someone who was
geographically positioned to help us out and located in the appropriate social stratum. And we needed to maintain security. We
couldn't let our target know he was being tracked, since this would
give away our client's intentions. In effect, what we needed to do

176

THE

INTELLIGENCE

EDGE

was access the gossip domain to see what was being emitted into
that sphere by the new CEO, and we needed a collector in place
who would also serve as a screen over our activitya fancy way of
saying we needed to hear the gossip and the country club was a
great place to listen if we could recruit a member.
Again, we asked our staff: Did anyone know anyone who lived
in that particular town in New Hampshire? Not surprisingly, no one
did. We extended our query to some of our outside professional
supports, whom we frequently used as extended parts of our network. A partner in our accountant's firm came from that town. We
explained to our accountant, under the seal of professional confidentiality, what we were after and why. We made it clear that we
were not seeking anything illegal or even vaguely unethical. All we
wanted was information that was already public, if unpublished, in
that town. We made it clear to him that we did not want to be connected with the query in any way, but we did want to know what the
word was in that country club about our man's intentions. We
wanted no confidentialities violated, no secrets stolen. We merely
wanted a report back on what was being said publicly.
Our accountant agreed to serve as go-betweenhe liked
being our accountant and this was a service that would enhance his
value to us. Alsoand this is importantthere is a mystique in
covert intelligence gathering that excites the Walter Mitty in many
people. It is something that can be taken advantage of. Anyway, he
spoke to his partner, explaining what was needed. We had asked
him not to tell his partner about us, but this was a fairly empty
request. First, his partner could probably infer which of their clients
being handled by our accountant would make such a weird request.
Second, our accountant's loyalty to his own firm probably transcended his loyalty to us.
As soon as you tell someone that something is a secret, their
brains go into overdrive trying, to figure out what the secret is.
In going to our accountant, we were making a conscious decision to accept a calculated risk of leak. Face it: Every time you take
a matter to another human being, you run the risk of some degree

When You Absolutely, Positively Have to Know

177

of betrayal, intended or otherwise. This is one of the great disadvantages of human intelligence. It is far easier to control security
breeches during technical intelligence projects than in human
intelligence. The very thing that gives human intelligence its
strengthcontext and nuancealso allows agents to see and
understand far more than you want them to or hope they do. It also
makes them unpredictable. Would our accountant tell his partner
who we were? Would the accountant guess? Would he guess who
our client was? Would he tell his wife? Would she tell her mother?
Would her mother tell his mother, who still lived in that town.
Would unexpected friendships or loyalties compromise our operation? Would he let all this slip unintentionallyor for personal
profit?
This was the hard part. We most emphatically did not want this
guy to know that we were investigating him. We were afraid of two
leaks. First, that anyone we approached in the club would tell him
what we were up to. Second, we were afraid that a leak would occur
during the course of our investigation, while questions were being
asked. We needed to proceed with maximum stealth. It was a risk we
hatedbut hadto take. At each stepthe query to our accountant, his query to his partner, and the partner's search for access to
country club gossipthe risk of leak and exposure grew.
The fewer the links, the more likely the operations will be
traced back to you. The more links, the more likely a leak will
occur somewhere.
In sensitive human intelligence matters, therefore, it is imperative to keep the number of links to the actual agent to a minimum
consistent with the need to shield yourself. One main reason is
security. The agent, the person actually gathering the information,
obviously must know what he's after. As with all collection systems,
context and nuance are essential. Transmitting that context to die
collector means transmitting it through each and every link. This
increases die knowledge embedded in the pipeline and increases
the damage caused by a leak. Thus, on the whole, the fewer links
the better.

178

THE

INTELLIGENCE

EDGE

Another reason for keeping links to a minimum involves the

reliability of transmissions. As a kid, you probably played a game
called telephone. Seven or eight kids would line up. Someone
would tell a simple story into the first kid's ear, and they would successively whisper it to each other. It was amazing to see what came
out at the other end, after retelling by seven or eight people. By the
time you were done, the story was unrecognizable. That is the problem of transmitting through too many links. In handling agents,
two-way communication is needed, which doubles the possibility of
transmission error, massively multiplying the possibilities of miscommunication. This means that, from a transmission-quality
standpoint, minimizing the number of links is essential.
From a security standpoint, however, it cuts both ways. On one
side, there is the risk of leaks and betrayal; on the other side, there
is the need to shield the questioner. In this particular project, we
decided that the priority was on securityit would be better to have
a distorted answer than to have the questions traced back to us.
That decision was situational. There was no principle involved
beyond the nature of the case,
In the end, this relatively simple query produced a query structure looking something like that shown on page 179.
Our accountant's partner contacted five people he knew from
his hometown. We asked him, through our accountant, not to ask
any questions at this point, nor even to ask about the guy we were
tracking. All we were interested in at this point was someone who
belonged to the country club. We really did not want someone who
knew our CEO intimately. Quite the contrary. An intimate friend
would almost certainly have alerted the CEO to our interestthe
worst outcome. Our reasoning was that given his prominence at the
country club, given the recent death of his father-in-law and his new
position, and given the fact that in addition to interest, there was no
perceived need on their side for security, there was an extremely
high probability that country club gossip would be both rich with
information and, on the whole, accurate. So we were obviously hoping to avoid his best friend. What we were looking for was an active
member of the clubsomeone who hung out, played golf, had
some drinks, and went to parties.

When You Absolutely. Positively Have to Know

179

PROBING THE C O U N T R Y CLUB

Our accountant's partner sequentially contacted people,

beginning with those he knew best and therefore provided the least
security risk. It took five calls before he found someone he was confident was well connected to the crowd. This guy became the agent.
We had now reached the critical point in the operation. The agent,
whom we never met and never would meet, had to be fed the question, and there was no way of being cute. He needed to know that
someone was interested in finding out how the new CEO felt about
the company. What were his plans? Had he sold his house? Had his

180

THE

INTELLIGENCE

EDGE

family moved? What had he been saying? Ideally, our accountant's

partner could have worked this into a casual phone conversation,
saying that he had heard that So-and-so had died and turn the matter into idle gossip. Had he found someone in the know in the first
few tries, it might have worked. Unfortunately, he was down to number five on the list before hitting pay dirt. This guy was someone he
never called casually, so it had to be turned into a purposeful call.
We were now facing a dilemma. What's in it for the guy he's
calling? Those involved to this pointus, our client, our accountant, and even the accountant's partnerall had financial motivations of one sort or another to pursue the matter. Our potential
agent had no such motivation. We could provide him with oneby
paying himbut a member of that country club would not be a
poor man. Everyone has a price, or so we hear, but the price for
turning a well-to-do man into a spy is usually substantially higher
than the project is worth. Moreover, in approaching someone with
a money offer, you already shape the relationship. It is extremely
difficult to offer someone money for spying and then, when he gets
insulted, ask him to do it for free. Our policy is simple: We almost
never offer money to a potential agent unless it is a completely
impersonal job. We try to find a plausible way to make it sound
like help for a friend and then let him ask for money if he's
going to.
Offering money is a big step. The deal is no longer casual, no
longer minor. It increases your own vulnerability.
The partner was dealing with an old friend from high school,
someone he hadn't seen in years. There was no use in faking sudden nostalgia. The key was to determine if the friend was on intimate terms with the CEO. If he was, then we instructed our
accountant to have him back off. If he knew him at a distance, we
asked him to explain that he, as an accountant, was involved with
some people who were interested in the status of the firm now that
the founder was deadsomething both true and not particularly
revealing. He was wondering if his friend had heard any gossip
about what the new CEO's plans were.

When You Absolutely, Positively Have to Know

181

The fat was now in the fire. But it was in the fire in such a way
that we probably weren't going to get splattered. There was no reason for the accountant's partner's friend to feel used, nothing to be
ashamed of, nothing illegal or unethical involved. The old friend
said that he would ask around. The accountant asked that the matter not be traced to him because he did not want to appear to be
stalking or harassing the guyplausible, reasonable, and surely
something a friend could do for an old friend. They talked about
family, reminisced, and swore that they had to get together soon.
We had our agent in the country club.
Now things were no longer in our control. This was not an
agent who could be controlled by us. We had to rely on him to actually ask around, ask the right questions, ask the questions with delicacy, understand the answers, and communicate them back to his
friend, our accountant's partner. But we had enough cutoffs
breaks between ourselves and die agentthat we had traded
minimal control for maximum distance. Fair trade, under the circumstances.
The old friend started to ask around. He had heard gossip but
had not paid much attention. In the information soup of the country club, this was not a matter of interest to him before. Out golfing
one Friday afternoon, he asked a golf partner if he knew the CEO
and if he knew what his plans were. His golf buddy said that he
didn't know him but he would ask around. No big deal. That night
at dinner, he asked another friend if he knew the story. He did. He
had heard that the CEO was extremely unhappy about having to
leave his job, because he felt he had a shot at becoming the CEO
therea much bigger company in his own area of interestbut
pressure from his wife and her family forced him to take the job.
But, he said, that was a few months ago, he might have adjusted to
the new situation. He'd ask around. His wife, hearing this, said that
she knew his sister well. She was going shopping with her next week
and would let him know what she found out.
Our agent now had two subagents without any direct knowledge working on the case, and a subagent with knowledge. One
subagent never panned out. He asked someone who didn't know
anything and gave up. The friend's wife asked the CEO's sister and

182

THE

INTELLIGENCE

EDGE

I When You Absolutely, Positively Have to Know

hit pay dirt. His dinner companion asked three people, two of
whom had information.
All of this came back to our agen t as follows: One person said
the CEO was pretty happy with his new position while two other
sources, including his sister, said that he was desperately unhappy.
The agent, without consulting us, dropped out the odd story and
gave us only the unnuanced tale: The CEO was unhappy, primarily
because the finances were a shambles and he was afraid that his reputation was going to go down with the company. He wanted out.
This came from the agent, through the accounting partner,
our accountant, to us, to our client. And it turned out that part of
it was true and part false. The CEO wanted out because he was
ambitious. But the finances of the company were not in shambles.
He was simply unhappy with his position and trying to build himself
up by bad-mouthing his father-in-law's managerial skills. But we had
our report.

[',
s"
'

|;
jj'

Problem: How did we know this report was accurate? Answer;

We didn't. We could have mounted another separate operation to
see what was confirmed, but we didn't have (he resources. Moreover, in dealing with gossip, the problem is that the soup itself
might be false, and reentering the soup would not have improved
the quality of data. We could have hired a private investigator, for
ridiculous amounts of money, but he would merely have done what
we did, or tipped our hand.
After all the work and effort, we needed to sit down and think
through whether the information we had was reliable. In the end,
it was our client's decision. It was his call because it was his risk. He
wanted us to be more certain than we could possibly be. We could
have faked a false certainty, but the fact was that there were too
many links, too many agents, too many of them collecting without
a real motivation to be accuratetoo much chance for error. But
the core report was that the man was tinhappy. A buyout was possible. An approach was made. A buyout did take place. We were
heroes.
Now, this scenario was obviously case-specific. That's the
nature of active intelligence. Unlike passive intelligence, where you
can pretty much go down a checklist, in active intelligence things

183

change with each case. Nevertheless, certain things are constant.

The most important is this: Know what you want and figure out
where you're likely to find it. In this case, we decided that a country club was the best place. The possibilities are endless, and
some are obvious: trade shows, which are like huge bowls of ice
cream begging to be eaten; charity functions with concentrations
from the relevant industry; bars frequented by company employees.
Virtually any persistent gathering of people is a potential intelligence gathering site, so long as you know what you are looking for.
Then there is the question of how to approach the site. Sometimes you can be blunt. We love calling a company, asking for a
given department, andjust asking a question. Cold calling works, if
only because the guy at the other end is taken by surpriseand it's
open and honest. Sometimes, though, you can't be open, honest, or
blunt, so calling and pretending to be a customer works. On rare
occasions, hiring an investigator to follow someone around works
but only if you're interested in who he's meeting with. Otherwise,
it's pretty pointless.
Knew what you want; have a good idea where to find it; find
the right person to go after it.

CONCLUSION: THE USES AND LIMITS OF

( A C T I V E INTELLIGENCE

'. The key to active intelligence is the clear identification of the infor| mation being sought and a rigorous definition of where that inforI mation is likely to be found. In our operation, we knew what we
*' wanted to know and identified a location where that information
was to be found. With that clearly in hand, we figured out a way to
get someone into position to hear what we needed to hear. Once we
Heard what we needed to know, we terminated the project.

When you've got what you want, ge hone while yoo're

still ahead.

184

THE

INTELLIGENCE

EDGE

This can often be the most important rule to follow, the most
important part of the operation. Open-ended active collection
operations can get very expensive and very hairy. In this operation
we were very careful to "stay on the sidewalk." By this we mean that
the sidewalk is public property. If, while walking down the street,
you pick up something on the ground or see something taking
place, even on private property, you are still clearly within the law.
Staying on the sidewalk means staying within the public domain.
Widely circulating gossip is within the public domain. Finding out
what was being said in a social circle is perfectly legal, even if it
was rigorously planned and carefully targeted. And it can be cost
effective, precisely because it ended when we found out what we
needed.
Long-term, unending operations, in our experience, have a
nasty and frequently unintended consequence of going off the sidewalkinto private property. Let's say that you want to find out more
about what a competitor is doing. A neat way of doing this is to have
someone answer one of his help-wanted ads. Initial interviews can
be very revealing. You're OK since they are publicly broadcast, and
hopefully the interviewee is not being asked to sign confidentiality
agreements, but is openly stating that he's happy where he is but
wants to hear more about what this company is doing. But let's
assume that he's called for another interview or that he's offered a
job. You're on a slippery slope filled with temptationto mix
metaphors. You now have the opportunity to plant an agent directly
in the belly of the beastthe monster competing with you. Is there
any situation that would justify this?
It might help to remember a famous case in the world of investigative reporting. Two "Prime Time Live" producers applied for
and accepted jobs with Food Lion, a food-handling company, on
the basis of false employment records and references, then proceeded to tape and film the inside workings of the company using
hidden cameras and microphones. What did they get? A good story.
What did it cost them? When discovered, they were sued for fraud,
trespass, and breach of loyalty for their falsification of records to
obtain the jobs. Note: They were not sued for libel, so the information they obtained was probably accurate. Maybe their case wasn't

When You Absolutely, Positively Have to Know

185

long term or unending, but it was certainly off the sidewalk. The
jury awarded Food Lion $5.5 million.
What are you going to get from using these methods? Your fantasy is drat you '11 get everything, but that's not likely, any more than
someone working for you is likely to get "everything" about you. Are
you going to get more than if you have someone hang out at the
coffeehouse where your competitor's employees go? Could the
"Prime Time Live" producers have accumulated the desired information any other way? Maybe not, but was it enough to justify the
cost and the risk? Well, if you are as vague as most people are about
what they are after, almost anything can be justified. But if you are
If as focused and specific as you need to be, then you'll probably realize that the risk isn't worth it.
Focus and specificity can save you from the worst thing in intelligence work: the unlikely outcome. Even with the best security you
can never tell how something will leak, how some extremely
unlikely coincidence will blow up in your face. We cherish the story
of one of our friends who lived in a small town in Pennsylvania, A
respectable fellow in most ways, he did have a lover. Once, on a business trip to Los Angeles, he arranged to take her along. He followed
security procedures in every way. She traveled on a different day on
a different airline. He checked into the hotel room only under his
name and she never answered the phone. Thousands of miles from
I home, they were anonymous, secure, safe. They went walking down
1 the street in Century City, hand in hand, very much in love. Sitting
I' on a bench, ten feet away from them, was his minister's wife, attends ing a convention for minister's wives in Century City. What were the
I odds of that happening? In our experience, too darn high.
Ongoing intelligence operations are going to blow up in your
face. Count on it. In a world governed by Murphy's Lawwhere
anything that can go wrong will go wrongextended and interminable intelligence operations are accidents waiting to happen.
The longer they go on, the more likely they'll go wrong. The
broader they are, the more likely they'll bump into something. You
I may think you don't care if people find out that you are spooking
around, but you do. By being known as a spook, people will throw
I up barriers to your most routine, passive, and semiactive questions.

186

THE

INTELLIGENCE

EDGE

Denied easy access to the public domain, you will find that, in the
long run, you will know less than you did before and at much
higher prices. Promiscuous active intelligence just doesn't pay. And
that goes triple for illegal active intelligencealso known as
espionage.

LUCKY GUESSES
The Art of Analysis

Analysis is where the rubber meets the road. The interpretation of

informationanalysisis the tissue that connects business intelligence with business operations. All of the intelligence gathering in
the world, no matter how subtle and effective, is merely the raw
material that allows the analyst to cook up his brew. Remember our
example of the first business intelligence operative? When Joseph
helped the Pharaoh corner the Middle Eastern grain markets he
was more of an analyst than an intelligence gatherer. After all, it was
his ability to interpret the Pharaoh's dreamsto analyze them
and produce an action plan that allowed him to be so effective.
Without this ability the operation would never have gotten off die
ground. The point of gathering tea leaves is to read them. Without
being able to read them, all the gathered tea leaves will constitute
nodiing more than a giant mess on the table.
The product the analyst produces is advice, and it can take
many formsfrom regular briefings to formal newsletters to web
sites with fancy interactives. No matter how complex the delivery
system becomes, the analyst serves his customerthe decision makers in the company. Now his analysis can be distributed to other
| employees, and even to customers or the entire industry. But the
Crocus of the analyst is, and must be, the corporate decision makers.
This is an important part of analysis that is frequendy forgotuntil the analyst is sacked. Business-intelligence analysis is not
187

188

THE

N T E Ll_ I G E N C E

EDGE

an academic study, a search for truth that will withstand the test of
the ages. The sole purpose of business analysis is to provide the specific customer in the specific company what he needs in order to
make decisions. Please notice that we said "needs" rather than
wants. That is and always will be the crisis of the intelligence analystas it is the crisis of the doctor, lawyer, or accountant. What the
customer needs is not always what he wants. But an intelligence analyst who doesn't have the confidence of his customer is useless, and
the one who wins that confidence by giving his boss only what he
wants to hear is equally useless.
The intelligence analyst is only as good as his customer: This
is where the ultimate responsibility rests. Unfortunately, as
lawyers hnow only too well, it is easier for the customer to
blame the analyst than himself. And that, as we say, is life.
Every businessman is also an analyst. One of his jobs is trying
to figure out what is going to happen next, and to position his company to take advantage of it. So one of the first decisions in creating
an intelligence system is to define the manager's role in the overall
process. In a small business, where the entrepreneur is already acting as CEO and janitor, he or she will continue in the role of analysthopefully with some new insights about that role and die
process of collecting intelligence. But in larger companies, even
though it is prudent to minimize the cost of intelligence by dualtasking collection with normal business functions, separating analysis from operations remains an important thing to consider. We have
already touched on some of the reasons for such a separation, as
well as some of the arguments against it. In order to go into the question in more depth, we need to think about what is involved in analysis and then consider the necessary functions and skills of an analyst.

THE A R T A N D C R A F T OF A N A L Y S I S

What do we mean by analysis? Does simply diinking about one's job

constitute analysis? In a general sense, it is obviously necessary ior

Lucky Guesses

189

every employee to spend time thinking about his or her jobGod

help the company where this does not happen! However, when we
talk about intelligence analysis, we mean something a bit different
than ordinary thought. It must be more self-conscious and selfaware than the ordinary, necessary process of thinking about doing
ajob. Intelligence analysis must have a different and much clearer
purpose, focus, and method.
The foremost purpose of analysis is to translate information
into a particular type of knowledge, which, in our introduction, we
called situational awareness. Situational awareness is a term we borrowed from fighter pilots, who live in a world in which the collection of information tends to outstrip the ability to absorb and
analyze it in the time frame in which decisions have to be made.
Fighter pilots are moving at near, or even above, the speed of sound
and dealing with other, potentially hostile, aircraft doing the same
thing. Sensors are gathering information about everything from
the temperature of the turbines to the location of friendly and
enemy fighters. The pilot must absorb all of this information and
make decisions. He can't spend a great deal of time reading each
dial, looking at each digital readout, or studying a radar screen. He
doesn't need any one piece of information nearly as much as he
needs an intimate, intuitive grasp of the entire situation around him
in one comprehensive gulpwhat psychologists call Gestalt. Military aircraft designers are spending a great deal of money trying to
solve the problem of "information overload."
Information overload is simply excessive information, information that has not and probably cannot be digested. A pilot cannot afford to spend too much time interpreting or analyzing
information. He needs to have information delivered to him as
knowledge, in a predigested form, so that he can, with minimal
effort and distraction, absorb it and turn it into situational awareness. Situational awareness, then, is a knowledge of the whole situation, constructed out of the pieces of information that are surging
toward him, and fused into a synthesis that is readily understandable and can provide the pilot with the knowledge needed to make
decisions in the next tenth of a second.
Given enough rime, the pilot is quite capable of understand-

190

THE

INTELLIGENCE

EDGE

ing every bit of information provided to him. Indeed, ii'he did have
time to study every bit of information coming into his cockpit, he
would undoubtedly have a superior situational awareness. But trying to do so would paralyze him, making it impossible for him to fly
and fight his aircraft. So he has become increasingly dependent on
systems that analyze the information for him, fusing it into a single,
comprehensible, real-time situational awareness. The push for virtual reality in the cockpit is an attempt to force computerized analysis to its limits, creating an automated analogue of reality, which
fuses all incoming information into one useful, if not fully true, picture of what is happening out there.
Effective decision making requires that situational awareness
be effortless. Achieving an effortless sense of reality takes a
lot of work.
The fighter pilot requires someone or something else to do his
analysis for him, since the decision point passes well before he can
absorb the material. Left to his own devices, the pilot is too far
behind the knowledge curve.
What has happened in the cockpit, we would argue, is also
happening in the executive suiteif not in quite as dramatic a fashion. The time needed to fuse information into a knowledge of the
situational awareness is expanding dramatically. Obviously, business
executives would be substantially better off if this were not the case.
The optimal situation is always one in which the decision maker is
intimately familiar with every morsel of information before acting
on it. However, what is optimal is not always possible, both in the
head office and in the cockpit.
Just as the pilot needs tools to fuse information into a comprehensive situational awareness, so does the businessman, who, in
a sense, lives in a much more complex, less orderly, and unpredictable environment. The variables that are matters of life and
death for the pilot are fairly well defined, more predictable, and
mosdy physical. While engaging a MiG fighter, it is unlikely that he
will be seriously effected by Japan's new import regulations.
The businessman, however, can be affected by virtually any-

Lucky Guesses

191

thing. Information on so many potentially important things are

constantly flowing into his business. Most of these things are irrelevant; some hint at wonderful opportunities; still others point to
deadly danger for his business. This is where analysis must extract
from this information soup those things that are of real significance
to the business, identify variables that are both changing and likely
to require decisions, and present them in an efficient manner. The
direction and nature of threats and opportunities for the businessman are highly unpredictable in location and type. They are also
rarely physical. Usually, a businessman can't spot it on a radar
screen; his screen is much more complex and subtle. The collection
process is of no value if it overloads the executivethe last thing he
needs is more data dumped on his desk. As with the fighter pilot,
excessive, undigested information can retard the decision-making
process. The executive and the fighter pilot both need intelligenceknowledge at the right point on the curverather than
more information.
The purpose of analysis in business intelligence is to present
the decision maker with a full and comprehensive awareness of
what is going on around him, in such a way that he can make a decision or request and receive additional, detailed information quickly
and efficiently. Some day, as with the fighter pilot, microprocessors
might exist that can produce this situational awareness. But today,
while microprocessor support might be available, it takes a human
being to turn information into knowledge. The analyst is the
microchip that processes and fuses the incoming information
quickly enough to alert the executive to impending shifts in his surroundings. The executive may have time to do this himself, or he
may hire someone to do it for him. This is where the focus of analysis and the role of the analyst overlap and need to be considered
together.

THE

ANALYST

The analyst needs to be comprehensive, but also focused, in his

work. That is a tall order. A soap manufacturer needs to know about

192

THE

INTELLIGENCE

EDGE

much more than the manufacturing of soap. Everything from the

latest fashions in Brazil to European Union environmental regulations to information on interest rates in Japan are of potential interest to him. Indeed, since the unanticipated is the most dangerous,
it is literally impossible for him to circumscribe his universe of information. That universe is usually circumscribed by available time
and resources; so inevitably, some event will take him by surprise.
Had he been able to anticipate it, he could have dealt with it far
more effectively. Context and nuance, as discussed in Chapter 3, are
indispensable if the significance and relationships of facts to each
other are to be recognized.
This is one of the central problems in going outside your own
organization to do intelligence work. A business analyst drawn from
a consulting firm can certainly carry out excellent research on a
detailed question, but he cannot provide the kind of situational
awareness required unless he is intimately familiar with the business
realities a decision maker is dealing with. A company having an
inventory problem is going to be watching short-term interest rates
much more carefully than a company whose concern is a new
emerging competitive technology. At the same time, the intelligence analyst must know what is of potential interest to the business,
so that he can provide timely warnings of dangers and opportunities.
The analyst must be given access to a company's strategic
thinking as well as to operational details. That is why, in one sense,
the ideal analyst would be the head of the business. Leaving aside
the important question of self-criticism, no one knowsor ought to
knowthe company better than its CEO. That means no one is a
better analyst. This is one of die great advantages small companies
have over large ones. While large companies have more room to
maneuver and greater economies of scale, small companies have
superior situational awareness, because strategy, operations, and
analysis are all located in the brains of one or a handful of people.
Second, the analyst must be intimately familiar with the types
of information available. He must know what information has to be
gathered and absorbed to provide baseline situational awareness
and what must be available to deal with specialized queries. He

Lucky Guesses

193

must be familiar with the strengths and weaknesses of all his

sources, know what he can ask for, where he can ask for it, and how
much trust he can place in it. The intelligence analyst must, therefore, be both a trusted member of the company's management
team, with ready access to the decision-making process as an
observer, and intimately familiar with the information flowing into
the company, or that might be brought in.
An analyst without access to the secrets of the executive suite
is like a doctor who can't examine the patient. All you'll get
from him are the bills.
The analyst must also be a magician, a soothsayer. Like Joseph,
the intelligence analyst is a fortune-teller. The ultimate value of
intelligence is prediction, which, after all, is the difference between
intelligence and journalism. A journalist tells you what has happened; the intelligence analyst tells you what is going to happen. If
an intelligence analyst can't tell you what will happen, then all you
need is a subscription to the Financial Times. Now, obviously, there
are very few fortune-tellers among us with the skill of Joseph, who
could turn the Pharaoh's neurotic nightmares into a business plan.
And this is the crisis that every intelligence analyst faces. Lacking
Joseph's skills, most analysts can't risk categorical declarations. The
guy who said that talkies were a passing fad and silent movies would
continue to reign probably had trouble getting anyone's attention
with his prediction that TV would never work.
In order to achieve even these modest goals, a good analyst
needs to have certain moral and intellectual qualities. The most
important of these are honesty and courage. The foundation of
intelligence analysis is a willingness to be wrong, wrong in two
senses. First, it is the willingness to change your mind when the facts
demand itnot to fall in love with your own analysis, not to let your
own ego get in the way of recognizing error. Second, it is the willingness to stand alone, against the crowd when necessary, even if it
turns out you were wrong.
The intelligence analyst must have an ego so large that he is
genuinely more concerned about what he thinks of himself than

194

THE

INTELLIGENCE

EDGE

what others think of him. This allows him to stand free of trends. It
also frees him from the need to be an iconoclastallowing him to
go with the trend when that seems reasonable to him. Honesty and
courage are the critical elements, along with deep-seated ambition.
There are many businessmen whose driving ambition is to be average, to do as well as everyone else. They would rather be wrong than
alone. The real intelligence analyst is different: He has to be indifferent to whether he is alone or in good company. He has to be
more obsessed with whether he is right than widi whether his boss
likes him.
Intelligence analysts are strange people. You may invite them
to the office party, hut it is doubtful that you want one
marrying your brother or sister.
Intellectually, the intelligence analyst is caught in a two-dimensional struggle. On one hand, situational awareness requires that he
be a generalist; on the other hand, mastering specifics requires that
he master narrow areas of expertise as well. That is one dimension
of the struggle. The other is a struggle between the discipline of
believing only what he knows to be true and the need to be intuitive, or guessing successfully at the parts he doesn't or can't know.
The intelligence analyst's mind must be synthetic: It is an artifact
that the analyst has to craft for himself, as if he were crafting any
other tool of the trade. It is also synthetic because its task is to synthesize, to pull together all the diverse strands that constitute the
situation.
The best analysts we know try very hard to be stupid and sirnpleminded. Sophisticated, well-connected people are bad analysts
because, being well connected, they start to think like the people
around them. Bright and sophisticated people believe they know a
great deal: They are intimately connected with highly placed
friends, they go to meetings, get invited to give talks. They are in the
know, so to speak. They know what Henry Kissinger and Alan
Greenspan are saying at cocktail parties. They attend conferences at
the Harvard Business School. They speak with partners at Goldman
Sachs. They even speak with people at Microsoft. They are the best
and the brightest.

Lucky Guesses

195

From Henry Kissinger, they learned in 1977 that Iran would be

a tremendous place to invest on the international scene. Alan
Greenspan told them in 1982 that the economy could not recover
unless the budget deficit was brought under control. At Harvard in
1990, they learned that Gorbachev was the man, and that Boris
Yeltsin was a loudmouth bumpkin. They also learned that the Soviet
Union, whatever its problems, would survive. At Goldman Sachs, in
1992, they heard how the Japanese economy would overtake the
United States. From Microsoft in 1992, they learned that the Internet was not something that had to be taken seriously.
The assumption that is frequently made is that because
these sophisticated, well-connected people are privy to the inner
thoughts of powerful decision makers, they know what is going to
happen and that their worldview, if properly understood, will create
a map of the future that will guide the analyst. However, being
sophisticated has more to do with blending in with a mind-set,
going with the crowdalbeit an exalted and sophisticated crowd
than it has with being right.
All those who predicted that Arab petrodollars would rule the
world in the twenty-first century were replaced by others who
believed that Japan would rule the twenty-first century; they were
then replaced by people who believed that China would rule the
twenty-first century. The best and the brightest were wrong. An
intelligence analyst following their thinking ought to be working on
die editorial page of the New York Times, and not for your company.
Analysts should not be afraid to stand alone. The danger of
casual assumptions of well-connected sophisticates is twofold: It is
not simply that conventional elite opinion and expectations are frequently wrong, it is also a fact that accepting basic assumptions from
anywhere limits the analyst's imagination and causes the analyst not
to question and examine issues that must be examined.
The history of national intelligence is littered with such universally accepted assumptions that proved to be utterly wrong. Just
consider one incredible error with tremendous consequences for
business. In 1973, the Arab oil embargo raised the price of oil dramatically and caused gasoline shortages in the West. Sophisticated
opinionranging from such places as the World Bank to Harvard
to the Club of Rome to the Kremlincombined with the growing

196

THE

INTELLIGENCE

EDGE

power of the Arab world and with basic assumptions about the ecology to reach a basic consensus: We were entering an epoch of permanent oil shortage. It was agreed that with demand growing,
spaceship earth (they really talked that way) had a fixed amount of
crude oil available. Therefore, the price of oil would continue to
rise indefinitely. Indeed, the argument went, this applied not only
to oil but to all sorts of primary commodities like food and ores.
The age of expansionary industrial economies was over.
It followed, therefore, diat investment in an industrial plant
put you on the short end of the stick. People producing cars were
at the mercy of people producing oil. It made sense to put your
money into extractive industry and farmland, and avoid investing in
industrial infrastructure. Since most undeveloped raw materials
were to be found in the third world, it made more sense to invest in
die third world than to invest in advanced industrial countries.
An analyst who uses the phrase "as everybody knows" aught
ta look for a new career, probably in the food-service industry
as a party planner.
So, during the 1970s, huge amounts of money were invested in
Mexican, Nigerian, and Indonesian oil and other commodities,
while the Arabs seemed about to gobble up die West's industries. If
you wanted to build a plant in the United States, you went to Riyadh
or Dubai.
The problem was that the consensus was wrong in a fundamental way. The assumption was that we had discovered all the oil
there was to discover. With the price of crude soaring, the world
went crazy hunting for oil. And they found it. By the early 1980s, the
price of oil, and of all other commodities, was plunging. Oil-tanker
owners and third-world refineries built on the assumption that oil
would be selling for over $40 a barrel and up to $100 a barrel went
bankrupt as the price of oil plunged. And third-world governments,
which had borrowed huge amounts on the assumption that the
loans would be paid back from oil and other mineral revenues,
defaulted on billions of dollars' worth of loans. People who bought
Iowa farmland in the 1970s were bankrupt in 1985.

Lucky Guesses

197

Anyone following the consensus in 1975 would have invested

in Mexican oil, Iowa farmland, and Filipino government bonds.
They certainly wouldn't have invested in Chrysler, which, as everyone knew, was finished, along with Detroit. And it would never have
occurred to anyone that what they ought to be looking at had nothing to do with the industrial/mineral relationship with which everyone was obsessed but instead should be looking at start-ups with
weird names like Intel, Apple, or Microsoft. The consensus in this
case was not only wrong, it didn't even understand the issue.
Most people tend to think in straight lines. They figure that if
China's economy grew by 8 percent for the past five years, then this
will undoubtedly continue into the future, and by 2020, China's
economy will be larger than the American economy. Or if Microsoft
continues to grow the way it is growing, by 2020, everyone in the
world will work for Microsoft. This is obviously nonsense. The fact
that China has grown as fast as it has means that it probably won't
continue to grow that fast. The same for Microsoft. Good analysts
always remember that what goes up must come down. They therefore tend to think in terms of curves rather than in terms of lines.
The hard part is predicting the shape of the curve. But convincing
most people in 1975 that the petrodollar was a passing blip would
have been hard enough. People who believe in curves tend to lead
lonely lives.
There are two types of analysts. The first type wants to be right.
The second type wants to have company. One of the hardest things
in analysis is saying things that no one else believes to be true. No
one can be an analyst who is at root a conformist. This is obviously
true in stock-market analysis, where being a contrarian is embedded
in the entire system, under the theory that if everyone believes the
| market will go up, it can't, since everyone will have already invested
' their money in anticipation of the move. That contrarian spirit
must be extended to aJl sorts of intelligence analysis. It is the only
way things can be rudilessly and rigorously analyzed.

198

THE

INTELLIGENCE

EDGE

METHODS OF A N A L Y S I S

We use something called zero-based reasoning in our work. Like

zero-based budgeting, this is an approach which says that every bit
of what we assume to be the case must be reexamined and justified
on a regular basis. In part this is a mind-set. It is the thing that is
most obviously true, the thing that everyone knows to be the case,
that must be examined most carefullyand that is something that
requires a powerful will not to believe. But it is also a matter of
method. In Chapter 3, we discussed developing an information
inventory. Here, we would suggest a careful, written inventor}' of
assumptions.
An assumption inventory simply means defining the basic
worldview with which you are dealing. In Chapter 5, we discussed
someone looking to buy a pager. What assumptions does he believe
as he begins his research? Arbitrarily, a list might include:
People will increasingly want to be reachable wherever
they go.
The price of cellular phones will remain dramatically higher
than pagers.
The FCC will continue to permit pager operations at the current level.
The marketing of pagers and of pager connectivity will continue to include small businesses.
Any or all of these might be true. Most experts in the field
would argue that they are in fact true and will remain true. But if
you put your hard-earned money into a business, it is no comfort
that you, along with the experts, were wrong. Your goal is to
be right.
Therefore, the beginning of any analytical work consists of
gathering together all of the information that has been found, and
trying to extract from it the common, stable assumptions that most
people appear to share. These assumptions shape the literature, the
research, the argumentseverything including many of the facts.

Lucky Guesses

199

The analyst begins by extracting these core assumptions, identifying

them, and making them the focus of the research. The most obvious, reasonable assumptions should worry you the most, for example: Do people really want to be reachable wherever they go? Are
they going to become tired of being reachable? We don't know,
but that is, after all, the core issue in an industry selling perpetual
availability.
People who don't constantly reexamine basic premises are
people who say things like: Why would anyone want a
computer at home? What would you do with it?
Zero-base analysis carries with it a cost. You are, in effect, continually reinventing the wheel. By taking nothing for granted, the
analyst winds up constandy reexamining everything. In essence, this
is what the analyst shouldbe doing. It's the nature of the job. But in
practice, this gums up the works tremendously. We recommend two
solutions. Systematically, periodic assumption reviews should be
carried out. Existentially, one part of the analyst's brain should
always be studying the assumptions.
One solution, used in the defense community as well as the
intelligence community, is the Red Team or B Team approach. On
a major project, or on an issue of ongoing importance, teams are
assembled outside the group that normally deals with the problem.
Their job is to take a look at everything the Blue Team or A Team
has been working onin particular the assumptionsand tear
them apart. Obviously, this is a pretty expensive way to approach the
problem. We suggest, for small businesses in particular, that the
approach should be periodic, rigorous reviews.
Part of the issue here is how the intelligence process is
approached. Analysis drives, or ought to drive, collection. That the
collector needs to be guided in his collection is obvious. But in
order to do that, the analyst must translate a business's needs into
information requirements. This requires the analyst to have an intimate understanding of the strategic problem facing the business. In
turn, that understanding will guide the system. Strategic understanding allows the analyst to prioritize. He knows what is important

200

THE

INTELLIGENCE

EDGE

and when it will be important. He can also focus particular tasks

into a coherent whole. Most important, the analyst will be in a position to examine the premises driving the strategy, revealing shortcomings and redefining the strategy as needed. As part of the
assumption inventory, we would strongly urge periodic strategic
reviews. Obviously, these are already being done regularly in
healthy businesses. Regularizing these reviews, turning them into
the equivalent of strategic audits, provides everyone with an opportunity to rethink the basic issues facing the business. More important, by linking the strategic review to the intelligence process,
there is an increased probability that the review will not be carried
out in an information vacuum. An intelligence-based strategic
review does not consist simply of reevaluating the assumptions
behind strategy, but on reexamining the information in hand and
collecting new information. Strategic review without intelligence
gathering frequently means reinforcing existing beliefs, which may
or may not continue to be valid. Intelligence-based strategic review
becomes a tool for evaluation in the context of new information.
The ongoing task of the intelligence analyst can be divided
into three parts:
Issue-driven projectsspecific projects designed to solve
immediate problems. Where shall we build a new facility?
Will the city council pass a new zoning ordinance that will
affect us? Would a new product be made obsolete in two
years because of new technologies? Each of these projects
shares certain characteristics: They are limited in scope and
finite in duration. In addition, while they all may require
complex analysis and massive information, they all produce
fairly specific and limited answers. The vast majority of business intelligence projects, although not necessarily the most
important ones, lake this form. They are by far the easiest to
analyze precisely because they are finite.
Ongoing analysis of the operating environmentexamination of the immediate competitive environment of a business.
This is the type most closely associated with "competitive
intelligence." The task here is to define the environment

Lucky Guesses

201

what companies must be monitored, what products, what

governmental bodies, and so on. What variables must be
monitoredsuch as prices, technology, particular people.
Having specified scope and timing, systems must be designed
to monitor at the frequency desired.
Tours of the horizonsearching for the unanticipated. This
is the most difficult, the most necessary, and the most
neglected aspect of business research. Most businesses operate over extended periods of time in fairly constrained operating environments. Competitiors, personalities, technologies
in this environment are fairly predictable and change occurs
incrementally. Sometimes, change crashes in from the outside. A new technology makes your particular product obsolete; a new personality goes on a buyout binge; a new
competitor enters the field. The difficulty with this sort of
challenge is its unpredictability. Who could predict that electric calculator companies would be forced out of business by
a chip invented for NASA. Who could predict that a new
environmental regulation would push the auto battery business against the wall? Who could predict that Rupert Murdoch would decide to enter the TV production business and
create an entire new network? Well, these were predictable if
you were looking in the right place. This is the analyst's most
demanding task: knowing what to be concerned about while
it is still a distant problem, and taking it seriously.
No one, including the CIA, has the resources to do completely
satisfactory work in all three areas. Therefore, the task of the analyst is to decide how much time and resources should be devoted to
each task. This has to be done by the analyst, because this hierarchy
can only be created by someone with a full sense of the problems
associated with both the business's position and the difficulties of
the collection process.
The problem facing the analyst is the tension between the
need to complete short-term, finite projections and the need to do
long-term monitoring. Murphy's Law, that work expands to fill
available time, needs to be amended:

202

THE

INTELLIGENCE

EDGE

Work with deadlines attached take priority over work wlthiut

deadlines. That means that people will flounder around with a
$10,000 project on a deadline while neglecting a potential
billion-dollar deal without a deadline.

The analyst has a particularly acute case of this problem. As

work piles up, his ability to think and read in the long range
declines. As long-range judgment declines, so does his ability to
evaluate short-term projects. Retail-site selection that ignores a new
set of potential planning ordinances or the arrival of a megastore in
the region is not likely to yield the necessary resultsprofits. Analysts must be ruthless in reserving a proportion of their time for
monitoring the immediate environment and even more ruthless in
making certain that a careful watch is kept on the horizon. There
are no hard-and-fast rules for how many resources ought to be allocated nor how much timeso an arbitrary rule ought to be 50 percent/40 percent/10 percent of time and resources, with some of
the best intellectual power reserved for the third. It is in this final
10 percent that the great dangers and opportunities lie. It is also the
most unruly and unpredictable realm; the realm that requires the
most judgment.

ISSUES FACING THE A N A L Y S T

Judgment is at the heart and craft of intelligence. Everyone else can

process things, but the analyst must go beyond the process, stop,
and judge the meaning of what has been found. There are two
problems at the root of the analyst's dilemma:
The analyst lives in a world of imperfect and incomplete
information. No matter how much work is put into gathering
the best available information, it will remain flawed. The analyst nevertheless must make the best practical judgment possible in spite of this inherent weakness. "I don't know" may be
honest, but it is hardly useful to a manager who must make a
decision.

Lucky Guesses

203

The analyst can never be certain that he has all the information available. By definition, it is impossible to know what you
don't know. And since what you don't know could very well
kill you, the analyst is always inclined to continue the search
for more information, always afraid that the next report will
contain information invalidating everything. Putting off
reporting out of fear of surprise is an occupational hazard.
The analyst is caught up in a continual interplay between
information gatherers and decision makers. Decision makers want
answers and they want correct answers. The analyst in turn puts
pressure on gatherers to collect more and better information. The
point at which the analyst has all the information he can reasonably
expect to collect arrives long before he has gathered all the information that it was cost effective to gather. But how can the analyst
know that he has reached this point?
We have a simple, if not always reliable, rule of thumb. Most
research refers to other research and almost all business information refers to other business information. We reach a point in our
work where we are familiar with all the references being made. The
bibliography in an article refers to other articles we've already read.
Most experts refer to other expertsexperts we contact refer us to
experts we already talked to. Reaching this point seems to happen
quite suddenly: One day we realize we have exhausted the accessible information.
It does not mean that there isn't anything else to know, nor
that this material might not be useful. But it does mean that the
common network of discourse has been exhausted. From here on
out, we are in no-man's-land, where even the experts with whom
we've been dealing haven't gone. There may be some important
information that hasn't been networked into the publications and
expertise of people, but it is unlikely. This is the analyst's decision
point: Move ahead, or let it slide. There is no science to this point
it is the art of analysis.
Part of the problem here is the extent to which an analyst will
rely on expertise. Do the experts in a field really know what is going
on in that field? A distinction must be made here between expertise

204

THE

INTELLIGENCE

EDGE

in a finite area of existing informationsuch as a technologyand

the broader area of prediction. A genuine expert on magnetic
media in computers is probably a reliable source on the boundaries
of the field. Get a few of these together and, between them, they
probably know everything going on in the field. That does not
mean that they can predict where the field will be in five years. So
distinguish between expertise in an existing area and expertise in
prediction. This is perhaps clearest in a field like economics. If you
were interested in a small, technical area of economics, tracking the
expert would do the job. But all of the experts in economics
together could not tell you where the global economy was going to
be in ten years.
Expertise does not necessarily extend to fortune-tellingthe
thing in which an analyst is most interested. Analysts who rely on
experts to predict future trends are being sophisticatedand
they'll get nailed. The forces that are shaping the economy are
thoroughly impersonal. The richest and most powerful financier
has little control over and minimal knowledge about where we will
be in the future. Here it is absolutely essential to capture the impersonal essence of the forces that are shaping the world.
The distinction between the sphere where personal expertise
is valid and the sphere where impersonal forces can explain what is
about to happen, is the primordial mud of intelligence analysis.
This is even true in predicting people's behavior. A given politician
or businessman may not be the best source to consult on what he
will be doing in five years. He may not have anticipated the forces
that will come to bear on him and he may not have insight into his
own probable reactions to those forces. President Richard Nixon
did not think he would have to resign, but a good intelligence analyst could have predicted the forces likely to develop over time and
predict his actionswell before he did.
Do you really think Hoffa knew he was going to be whacked?
The insider knows what he hopes will happen. Sometimes
it does.
Knowing what people will do before they themselves know it is,
after all, what intelligence analysis is all about. Lest this appear to

Lucky Guesses

205

be impossible or absurd, let's consider a group of intelligence analysts who already do this on a daily basis: financial analysts predicting which stocks will go up or down.

FINANCIAL. A N A L Y S T S A S
INTELLIGENCE A N A L Y S T S

The financial analyst is a natural generalist who derives his vision of

the future by creating a stylized construct of the present, consisting
of financial statements, press releases, gossip, briefings, and the
like. Financial analysts who survey the broad markets for investors
are forced to take extremely wide, synthesizing views of reality. Even
here there is pressure to specialize in particular industries, but
good financial analysts look at their specialty industries from very
broad perspectives. Financial analysts are intelligence analysts, serving a specialized but extremely large segment of businessthe
financial markets. A significant source of information about any
company listed on pretty well any stock exchange is the reports produced for security houses and stockbrokers by their teams of financial analysts. Ostensibly, these reports are produced for the benefit
of major clients of brokerage firms so that they may decide if they
wish to invest in a corporation. But most stockbrokers make their
reports available to the serious media, which means newspapers like
the Financial Times and the Wall Street Journal; financial news agencies like AFX, Bloomberg, and Reuters, or business television networks like CNBC. The reports can usually be located through the
archives of these organizations and obtained directly from the
brokers.
The principal objective of the financial analyst is to assess the
value of the companies he studies, and to assess future potential
profits and share valuebasically to try to predict the future. To get
this kind of price-sensitive information he needs to probe deeply,
and get as much inside information as possible. So a successful anaL lyst will spend part of his time in the office poring over company
accounts, reading official filings to stock exchanges and bodies like
the Securities and Exchange Commission (SEC), and teasing out
the numbers. But then the analyst will want to visit the company,

206

THE

INTELLIGENCE

EDGE

meet as many people as possible, and also meet the company's competitors. The besl analysts will also want to visit at least some of the
company's operations or factories to make an assessment of their
efficiency. Herein lies the danger. The financial analyst has to make
sure he is not suborned by a new breed of flack, known as the investment relations consultant. The job of this person is to put a favorable spin on a corporation's financials, to try and ensure that the
analyst only meets the "right" people, and to keep the share price
as high as possible.
Investor relations managers tend to have generous expense
accounts. They are at their most active in Wall Street and in the
financial capitals of Europe, and they spend considerable energy
and charm in courting the analysts. This is the financial analyst's
equivalent of the sophisticated, well-connected source. Both those
close to the decision makers and the investor relations manager
types have a definite personal interest in presenting the best picture
about their organization or country.
Financial analysts are both role models and sources. But
beware. Contrary to their self-image, they are still human
and fall in love with their own advice.
The lesson here for those gathering business intelligence
about subsidiary companies or overseas operations that are secretive about their activities is to look at analyst's reports on their
major business partners, especially if these partners are based
in Europe or the United States, where analysts are most active
and where reporting requirements from the share markets are
rigorous.

PROJECTING THE FUTURE

In any attempt to project the future, the key question that must be
asked is: What will determine the future? Answering that question
will, in turn, determine what information will be required to get
a sense of the future. We can divide the factors determining

Lucky Guesses

207

the future course of events into two categories: personal and

impersonal.
There are events that basically depend on the will of people.
For example, whether this person or that person will be hired for a
job depends to a great degree on the will of the hirer. Whether a
new product will or won't be invented depends on what is going on
inside the mind of the inventor, investors, and so on. Whether a
particular company will attempt to buy out another company is, at
least in part, a matter of willand therefore you need to know what
certain people are thinking in order to predict the future.
There are other matters that really don't depend on will, that
are highly impersonal matters. For example, whether the stock market will rise next year is not in the hands of any one man or even
group of men-paranoid fantasies not withstanding. The path the
market will take depends on a wide variety of impersonal forces. We
can and do debate which forces are more or less important
indeed, that debate and the different perceptions and theories is
one of the forces that shapes the marketbut you cannot predict
the course of the stock market by having a long talk with any one
person. If it can be predicted at all, its course will be determined by
very public forces, data about which is widely available. In business,
a large part of the future is determined by market forces that, by
definition, are not under anyone's control.
The first task that any analyst has, the foundation of all of his
work, is to determine the extent to which his company's situation
and future are defined by personal and impersonal forces. Con[ sider: You are the owner of an independent but prosperous bookstore in a small town. You afe aware of a major trend in your
| industrygiant bookstores. You are doing pretty well but doubt
that you can survive if a giant moves into your area. Your lease is up
for a two-year renewal. If a giant is going to move in, it would be
rational for you to close operations now. If a giant is not going to
move in, then you certainly should sign up. Question: What will the
giants do?
At first glance, this is a very personal decision. There is obvif ously a team in place at each of the giants that has identified the
I markets into which they would like to move, and there are clearly

208

THE

INTELLIGENCE

EDGE

plans in existence for at least the coming year, and probably more,
for new openings. So, at first glance, an analyst would decide that
this project is dependent on personal inclinations and focus on
penetrating the planning cell, which may or may not be possible.
If you look at the matter a little more carefully, however, you
can see that there are two types of impersonal forces at work. First,
while planners have the power to select, the selections are clearly
going to be made not arbitrarily but by some predefined criteria.
Now, you can try to steal the document that contains those criteria,
or you can reverse engineer the document. You know where the
giants have opened stores in the past and you have an awful lot of
statistics available on those locations. You can take a look at where
they have gone and deduce ihe criteria they must be looking at.
Then it is a matter of looking at your own market and seeing if it fits
those criteria. This isn't foolproof, and, obviously, you cannot
absolutely predict the future, but past performance is a pretty good
indicator in this case.
There is a second set of impersonal categories you might look
at. The giants have been expanding heavily. This fact alone does not
mean that they will continue to expand at the same rate. Thev
might become more or less aggressive. To a very large extent, this
will not depend on them. More precisely, all businesses live in a
world of constraints, and they make decisions within the context of
those constraints. What does the company's balance sheet look like?
Expansion requires borrowing. How much more can they borrow?
They've had a massive expansion. Is their management holding up
or are they having problems? What does die retail book environment look like? Are computer games, the Internet, and other things
cutting into the market, ruining some of the projections they were
operating under? Are they going to be able to carry out any more
expansions?
In other words, an analyst, trying to figure out what all of the
information means, can go after the intentions or the capabilities of
the subject. This is the oldest dilemma in intelligence analysis and
one that will never be solved. If you focus on capabilities, you find
out what the other side can do, not what they will do. If you focus
on intentions, you are focusing on what they would like to do, not

Lucky Guesses

209

what they can do. Obviously, the answer is to focus on both. Still,
even in this simple case, you can get tripped up. Here are some
possibilities:
The giant's management is ignorant of its own limitations
and overexpands. Yes, they wind up in Chapter XI eighteen
months from now, but not before they expand into your market, wrecking your bookstore. The argument to focus on
capabilities assumes that your subject reads his capabilities
the same way you do. This is, then, a case where management's personal defect meant that they would be misreading
constraints and acting in a less than optimal fashion.
The giant's management has concluded that rather than
putting all their eggs in one basket, they would rein in their
bookstore expansion and diversify in sushi bars. This is a personal decision based on impersonal analysis. However, the
analyst's own impersonal analysis wasn't extended to include
the entire universe of optionsit never occurred to him to
study the relative rate of return on investments in bookstores
versus sushi bars. The bookstore closed and no giant bookstore ever showed up. We have here a case where personal
information controlled the area of impersonal knowledge
that had to be mastered.
The giant has generated more accurate data than the analyst
has available and is aware of that fact. The analyst's data
shows that the community is unsuited for a giant bookstore.
However, the data the analyst is working from is drawn from
a census area that is defined too narrowly. With better data
showing that shopping patterns in the region justify a much
broader geography, they have decided to move into the
region. So this is a case where the best available data, defined
by the depth of the business's pockets, simply isn't good
enough.
The giant has been growing at a rate of 20 percent increases
in square footage for three years. Everyone inside and outside die company expects that to continue indefinitely. The

210

THE

INTELLIGENCE

EDGE

analyst expects it to continue. Everyone forgets that no retail

sector has ever increased by that rate for more than three
years (not true, but we're writing this book). Sloppy thinking
on all sides, yours included, decides to draw a straight line
instead of the inevitable curve. The store is announced,
ground is broken, but it is never completed due to a massive
slowdown in the industry. So here was a case where everyone
can agree on something and it still won't be the way things
turn out.
We could go on. The point we are making is that the interplay
between personal and impersonal forces, choice and constraint
ultimately the interplay between business and its environment
cannot be tracked with scientific precision. In a world filled with
unintended consequences, unexpected by-products, knowing
whether to look at a company's plans and internal thinking or at its
external reality is a matter of art rather than science.
In general, our experience has been that personal contacts are
too overrated in generating information. This runs counter to the
experience of most businessmen, but that is because for most businessmen, the core experience of doing business is sellingthemselves, their product, or their company. And in selling, nothing can
replace the personal relationship. There is a tendency to extrapolate from what you know, to reason by analogy. In selling, the personal relationship is the key. It seems to follow that, in intelligence,
these personal relationships will give you a window into the mind of
your subject, which in turn tells you what he is going to do, if not
precisely what will happen.
We all have oor reasons. Any insider has his reasons for
talking to you. Those reasons affect truthfulness. Use your
contacts. Don't worship them.
Consider what happened to Western businessmen in Iran during the reign of the shah. Everyone doing business in Iran had close
contacts with Iranians, many of whom were relatives of the shah,
friends of relatives of the shah, and so forth. Every one of these FOS

Lucky Guesses

211

(friends of the shah) advised their Western contacts to stay in

Iraneverything was fine. Then the shah fell and these businesses
were wiped out. How could all these contacts have been so wrong?
Well, think about it. First, what makes you think that an FOS
knows if there's going to be a revolution? If an American were
approached in 1997 and asked who would be president in 2001....
well, he could have an opinion, but just because he was rich and
famous and an FOB (friend of Bill), how much would that count
toward his ability to predict who will be president? What makes us
think that a well-connected Iranian really knows what's going to
take place?
Second, and more important, what makes you think that he'll
tell you even if he knew? A friend of the shah makes his living by
being a friend of the shahand by getting you to do business with
him in Iran. If he were to advise you that the shah was going to fall,
would you invest in Iran? He wouldn't make money that way. Moreover, if the shah falls, then you can't make money being his friend,
right? So you aren't likely to tell your Western "friends," who keep
you rich, to stay out. Quite the contrary, you'll urge him to come on
in before the shah falls.
There are some things nobody can know; just because someone
is related to someone powerful doesn't mean they can predict
the future. LOOK whet happened to businessmen who were
close personal friends with the shah himself? Better still, leok
whet happened to the shah!
Listening to well-placed individuals is useful in giving you a
sense of what they are thinking. But many businessmen, who like to
be "in the know," get hooked on the idea that they've got a contact
who is telling him what is really going on. The only thing he is
telling you is what he thinks is going on, and that only if it's in his
interest to do so.
There is nothing wrong with using personal relationships to
gather additional information, so long as you don't fall into the fallacy that these insights will be definitive. It is absolutely vital to
remember that (a) people lie, (b) people frequently don't know

212

THE

INTELLIGENCE

EDGE

their own minds and can't predict what they themselves will do,
(c) the best-laid plans of mice and men..., and (d) some people are
idiotseven CEOs of major companies.
For all these reasons and more, we are major advocates of
impersonal studies of capabilities. While never definitive, they tell
us more about what people can do than about the fantasies in
which people indulge. And diey give us a sense of the limitations
that are imposed on people by reality.
In the end, it comes down to this. If having lunch with a CEO
would tell you what his company's stock is going to do, we'd all be
rich. CEOs themselves don't know what is going to happen to dieir
stock or even their industry. If they did, none of them would ever
be fired. Since they don't know, talking to them is useful only if they
are treated as indicatorsfrequently as contrarians.
Analysis is the art of the probable: What will probably happen
and when will it happen? It is not a game of certainties. Neither
being a CEO's best friend nor spotting a pattern repeating itself for
the eighty-third time in history guarantees that, on the eightyfourth time, you will be able to figure out what you are doing. The
very best you can achieve is something the trade calls the weight of
evidence. When everything is laid side by side, every fact is culled,
and you've sent out for all the information you need and have gotten it, the best you can do is take a walk, weigh everything, see
where most of the evidence lies, and trymore often than notto
be right. If you can be right, at least more often than others, you will
have done your business a great service and justified your existence
and salary.

CONCLUSION

The art of analysis requires detachment and disinterest. It requires

honest, utter self-confidence, and a total lack of ambition. What is
needed and what we have are, of course, two different things. But
some things are indispensable. One is that the analyst have access
to the executive suiteand to the entire companyso that he has
the context and nuance needed to do his work. Another is that he

Lucky Guesses

213

be respected enough to be listened to. In some ways, the second is

harder to achieve than the first.
This is whv there is a strong argument to be made that the
CEO ought to be his own intelligence analystat least he'll listen
to himself, and if he fails, he'll have no one else to blame. That is
attractive, no doubt about it. But it misses the practical realities.
First, the sheer quantity of material that must be assimilated makes
analysis a full-time job. The CEO will just miss too much of the
unexpected because he is too busy doing other things. Second, the
CEO needs a critic, a B Team.
During the 1980s, there was tremendous debate in the U.S.
intelligence community about the level of Soviet spending on
defense. The CIA had an official estimate but it was controversial.
President Reagan appointed a group of outsidersa B Teamto
take a look at the CIA's information and see what conclusion they
drew from it. The conclusion was quite different, and an extremely
healthy and edifying debate was kicked off.
Businesses desperately need B Teams, particularly those with
strong, charismatic leaders. This does not apply only to large companies but to small ones as well. Someone needs to occasionally
impose a reality check on the CEO. Reality checking is the specialty
of the intelligence analyst, perhaps his most precious gift to the
company. That is why we recommend, wherever possible, that the
role of intelligence analyst be carried out by someone other than
the CEO. Let us further suggest that the analyst be independently
wealthy, doing this work as a hobby. It will make those Friday afternoon meetings, where he tells the executive committee that an
Indian company is about to put out a product identical to their
Product Xthe savior of the company and brainchild of the
CEOand that it will sell for 35 percent of the price Product X will
sell for when it hits the market next year.

Taking Cover

215

When Phelps told this to a congressional committee in 1992

he was stopped in his tracks by the chairman, who asked him to
repeat the sentence for the record. Phelps replied:

T A K I N G COVER
Counterintelligence

Security and Counterintelligence are not, properly speaking, part of

the intelligence suite- But propriety aside, it is obviously a matter we
need to address. This is particularly the case when you realize how
revolutions in digitizing information have increased your own
access to information about the worldand about competitors. We
have been discussing all the ways you can access information. The
other side of that coin is the equivalent ease with which others can
access information about you. Now, some of the threats are very old:
disloyal employees, careless and loose talk, physical break-ins, and
the like. But some of the threats are radically new and unprecedented. You are used to locking your doors physically. But new technologies have created new doors, and you have to keep those
locked as well. Even more important, you need to keep critical
information from wafting out through unseen cracksto be sniffed
at by competitors as cagey in the intelligence game as you've
become.
Securing a company's secrets is expensive, but so is losing
those secrets. IBM is just one American company to claim to have
lost billions of dollars through the theft of its proprietary information. Marshall C. Phelps, Jr., the vice president for commercial and
industry relations, says that "the theft of corporate proprietary
assets has occurred from many quarters: competitors, governments
seeking to bolster national industry champions, even employees
we have been the target of each."
214

These activities have resulted in losses to IBM in the billions. Settlement agreements, jtidicial mandates, and
commitments to third parties preclude us from discussing the specifics of many of these cases, which
include the theft of personal computer-related technologies, theft of trade secrets critical to our mainframe systems, theft of information by governments for their own
use and by their indigenous hardware and software suppliers, and counterfeiting of our trademark on hardware
and software.
One of the most critical IBM proprietary elements
of our PC technology is software known as the Basic
Input/Output System, or BIOS. The BIOS controls key
output operations, such as interactions between the computer and diskette drives, fixed disk drives and the keyboard. This copyrighted software has been deliberately
and repetitively misappropriated by many domestic and
foreign companies seeking to manufacture inexpensive
copiesor clonesof the IBM PC.
Confirmation that Russia has been one of the countries stealing proprietary information came in testimony to the same committee from Stanislav Levchenko, a former KGB intelligence officer
from Moscow whose previous jobs included planting forgeries of
foreign government documents. He said that far from the end of
the Cold War meaning the end of espionage, there was a new target: American industrial corporations:
The remnants of the secret police are still there [F]or
decades political intelligence was the number one priority. Now high-tech intelligence and economic intelligence
is the most important priority of die headquarters and of
the residences abroad. The economic situation of Russia

216

THE

I N T E L L I G E N C E

EDGE

and the surrounding states is severe. They do not have

the resources nor the time for expensive research and
development efforts necessary to compete in international markets. To survive, they will steal the proprietary
secrets of foreign companies. It will be naive to hope that
President Yeltsin will decide to cut Russian intelligence
substantially. In addition, practically all the new countries, the former republics of the USSR, can be expected
to conduct intelligence on their own with U.S. technology as a prime target. Allegedly the Ukraine already has
hundreds of intelligence officers and is training the
younger generation. The intelligence services of the former republics will almost assuredly coordinate dieir
work. An important part of it will be industrial espionage.
The former Soviet Union was by no means the only villain in
the theft of American corporate secrets. After the hearing, the
chairman, Congressman Jack Brooks of Texas, said testimonies
from U.S. industry representatives and government witnesses had
"painted a dark and sinister picture of just how foreign economic
espionage activities by countries such as France and Japan have cost
American companies billions of dollars and hurt U.S. competitive-

ness.
Bavernments remain the first threat to your information.
Former governments are the second.
Nor, as Microsoft's Dr. Nathan Myhrvold reminded us, is die
issue only "about super computers for Uncle Sam, or paranoid
management information systems personnel at a bank."
It goes far beyond the issue of foreign economic espionage. Instead it ultimately is about every PC. It is about
doctors protecting the confidentiality of medical records
that they store in computers radier than on paper. It is
about preparing tax forms on a home computer and then
transmitting them electronically to the IRSwithout fear

Taking Cover

217

that the government or a hacker may obtain access to

your financial records or any other information you have
stored in your computer.
Myhrvold joined other leaders of the computer industry in
calling for a strengthening of encryption, and for the U.S. government to relax restrictions on the export of powerful encryption software. Encryption is based on a simple idea that can be traced back
to antiquity: transforming a message into a form that cannot be
understood by anyone without access to the encryption formula.
Julius Caesar was known to have used a crude form of encryption in
which he changed every letter in a message to the one that
occurred three letters later in the alphabet. Thus "attack at dawn"
became "dwwdfj dw gdzq."
We have been talking about the migration of intelligence techniques from the national security arena to the business world.
Along with the migration, we are seeing the movement of espionage techniques as well. That means, obviously, that national security techniques for counterintelligence and security will migrate as
well. In the end, diis will pose a substantial cost to business. Securing the knowledge that is the underlying capital of modern industrial society will not be an easy task. The very digitization of data
that has proven such a boon to mining data will become the
Achilles' heel of business,
At this moment, the threat to business still comes from the
national intelligence agencies. French and Chinese intelligence services are reputed to be particularly aggressive in trying to get hold
of strategic technologies. Recently, several members of the CIA station in Paris were expelled by the French, officially for engaging in
industrial espionage. Whether the charge was true or merely retaliation against similar charges made by the United States against
French intelligence is less relevant than the obvious fact that
national intelligence agencies, looking for a role beyond the politicomilitary sphere in which they operated during the Cold War,
have focused on accessing key emerging technologies in foreign
countries.
In a sense, this provides a degree of security for most busi-

218

THE

INTELLIGENCE

EDGE

nesses. Consider this: A highly centralized intelligence agency, controlled by a government, is going to focus on matters of the highest
national priority. Now, even if that no longer means focusing on
plans for weapons systems, it does mean focusing on technologies
critically important to the nation as a whole in its international competition. The United States and France are locked in a bitter competition in the passenger aircraft industry. That means that French
and American intelligence, tasked with supporting their respective
aircraft industries, will focus their attention there. Relatively little
will be left over for the countless other secrets in the countless
other industries.
National economic espionage will therefore focus on big-ticket
items, simply because that is the only cost-effective way to operate,
and because the political and bureaucratic realities dictate that sort
of approach. At this time, therefore, the focus of industrial espionage, from the state level, is on the top of the pyramid, and people operating in less salient, less exposed industries are safe from
the concerted and effective intentions of professional espionage
organizations. But they should not feel comforted by this.
The intelligence industry, worldwide, is experiencing a massive downturn. Two factors have come together to create a massive
economic crisis and layoffs. One is the general tendency to downsize governments worldwide. The other is that the end of the Cold
War and growing peace in the Middle East have thrown countless
sophisticated, top-of-the-line spooks out on the streets. Former KGB
agents, agents formerly employed by Eastern European Communist
states, now frequently outcasts at home; retired Israeli military and
foreign intelligence officials; CIA and DIA agents taking early
retirement or being RIFed (reduction in force), have created an
army of unemployed spooks. All of them have one thing to sell
their expertise.
And in most cases, their expertise is not in the modern areas
of passive intelligence in the electronic domain but in the older,
Cold War crafts of breaking and entering, wiretapping, tailing,
videotaping, and so on. Indeed, those who specialized in computer
espionage, information management and analysis, and such are the
ones most likely to remain employed. They are needed far more

Taking Cover

219

than those with older skills. So, since 1991, the world has had
hordes of nasties released on it. And all of them are looking
for work.
If a Ukrainian national conn* ta you with a deal and he speaks
perfect English, and his resume* says that he worked for the
Soviet Ministry of Agriculture as agricultural attache in Cairn,
you cauld do worse than pass on the deal. A lot worse.
We were involved in a couple of cases recently pointing in this
direction. In one, a Ukrainian national approached a client looking
to purchase some technology. It turned out that he was less interested in purchasing it than in stealing it. In another, similar case,
some Hungarians were looking to sell some chemicals in the West.
It turned out that they were dealing on die come: If they got an
order, they'd then go out and steal it. In both cases, we were clearly
dealing with former spooks, out of a job, trying to husde a living for
diemselves in the New World Order.
In many, if not most, cases these people live on the edgeor
over the edgeof the law. They work with and frequendy become
part of organized crime. In a sense, they are not nearly as dangerous as they appear, simply because their approach is not particularly
subtle; they have become thieves and extortionists more than intelligence agents. The threat here is less industrial espionage than the
increasing threat of organized crime on a global basis. The downsizing of intelligence services coupled with massive social instability
in the former Communist world has generated an entirely new,
increasingly global organized crime system. Less integrated into the
culture of international business than the old Italian Mafia has
become, they pose real, physical danger. But it is simple criminality
more than espionage.
The average businessman rarely faces a threat from a national
intelligence agency, and most mobsters are less interested in stealing intellectual property than in making you an offer you can't
refuse. The real threat that most businessmen face comes from the
same forces that have given them, in our opinion, their tremendous
intelligence opportunity. Digitization of information has allowed

220

TM

INTELLIGENCE

EDGE

them to tap into information resources previously unimaginable.

That same digitization inside their company opens the door not
only to intentional theft but, even more serious, the unintended
emission of important information generated by the business into
the public domain.

SECURITY

A N D SECRECY

In thinking about security and counterintelligence, it is important

to begin with Napoleon's famous dictum: He who would defend
everything, defends nothing. There are two important sides to this
truth. The first, more obvious one, is the simple recognition that no
one has the resources to genuinely secure everything he holds.
Security has a price, and not everything you possess is worth that
price. Someone who tries to secure everything equally will,
inevitably, be allocating equal resources to protecting things that
have no value to protecting things on which the future of the company rests. That sort of policy obviously makes no sense. Somehow,
a hierarchy must be created that frankly recognizes the relative
importance of information.
There is another, somewhat more subtle and complex side to
this as well. The foundation of information security is secrecy. The
more people who know a piece of information, the more likely that
information is to leak. That's obvious. And the solution would
appear to be equally obvious: Make sure that as few people as possible are familiar with critical informationthe old "need to know"
principle. Now, in many cases, need to know works, but it also has
costs and risks.
Secrecy is extremely expensive. One cost, obviously, is the cost
of securing information physically. But the deeper cost of secrecy is
inefficiency. In many large companies, research and development
projects, client lists, profit and loss statements are closely held. As a
result, one division might be developing at substantial cost a gadget
that another division had already developed but not yet released;
one division might be spending large amounts of money develop
ing contacts with a potential customer that is already a loyal cus-

Taking Cover

221

tomer of another division; one division might be undertaking a project identical to one another division undertookwith disastrous
results.
Secrecy protects information by limiting information's mobility. Sometimes it's simply confined to use by a few people. In other,
more extreme cases, it's broken up, compartmentalized, as the jargon would have it. In compartmentalization, the information itself
is broken up into pieces so that even those with access don't get to
see the whole. These moves work. They decrease die chance that
anyone outside die company will find out what is going on inside.
Secrecy also increases the probability that people inside the company don't find out what's going on.
Secrecy and security are cosdy. The major cost is inefficiency.
Duplicated effort, insufficiendy exploited capabilities, and lack of
synergy in carefully compartmentalized projects dramatically
increase costs for businesses. Part of the cost is the sheer expense of
physical security systems, from those on computers to those on
doors. Part of the expense is timetime spent accessing information. And part of die expense is the cost of failing to squeeze the last
ounce of utility out of a piece of information. The classic example
of all of fhis was the old Soviet Union. It had superb security yet
collapsed into inefficiency. The right hand didn't know what
the left hand was doing, because they weren't cleared to see the
information. The United States leaked like a sieve. Yet connections
were made between bits of information that generated entire
industriesprecisely because the information was not compartmentalized.
Secrecy and security carry with them another danger. When
information is held tightly, that means that fewer people know
what's going on. It also, therefore, means that those fewer people
are in a much more powerful position both within that organization
and in being able to betray the organization. By concentrating
information at certain pointsnot necessarily at the top of the
pyramid, with the chairman or even the CEO, but at the top of the
functional pyramid, in the hands of those individuals controlling
operationssecrecy increases the cost of security failures.
The digitization of information increases the possibility of

222

THE

INTELLIGENCE

EDGE

leakage simply because the efficient packaging of information

makes transfers easier. Consider the now classic case of GM, Volkswagen, and Jose Ignacio Lope? de Arriortua, whose defection from
the world's largest auto manufacturer, General Motors, to Europe's
top carmaker, Volkswagen, paved the way for the most bitter public
wrangle over industrial espionage in a generation. The row
involved criminal and civil court proceedings and politicians, diplomats, and leading businessmen in both Europe and the United
States.
Lopez, a devout Gatholic who often publicly carries a rosary,
was born in the Basque region of northern Spain and pursued a
career in purchasing, rising to the position of head of purchasing
for General Motors's European subsidiary, Adam Opel. There he
had won accolades for the way he had revolutionized the car producer's relationship with suppliers, allowing Opel to cut costs. As a
ranking and highly rated GM executive, Lopez was fully involved in
the corporation's plans for the future.
One of these came to be known as Project Xhardly an original description in a world where thousands of corporations have
concepts or ideas carrying this namewhich involved the creation
of a new car production plant in the Basque country. Project X
called for the techniques that Lopez had instituted to be fully developed in the new plant. The regional government was keen to bring
new industry to an area of high unemployment, and wanted to
emulate Ford's success a decade earlier when it had built a successful facility in Valencia, another area of Spain. The Basques offered
GM strong financial incentives to build the plant in their state;
after serious consideration the Detroit-based company turned the
idea down.
What happened in the ensuing months is a matter for speculation and the subject of $1 billion in lawsuits still pending. What is
known is that on January 1993, in a hotel in the German city of
Hannover, well away from Opel's headquarters in Riisselsheim or
from Volkswagen's home base in Wolfsburg, Lope/, sat down to
breakfast with Ferdinand Piech, who had not long before taken
over as chairman of VW. Also present was Gerhard Schroder, a former German foreign minister who had become the premier of the

Taking Cover

223

state of Lower Saxony. At fifty-nine, Lopez could have been contemplating a peaceful and lucrative retirement. Instead, as the
breakfast ended, he was handed a contract. According to The European newspaper, there was a blank space left for Lopez to enter his
own salary.
Within days, Lopez and three GM colleagues were established
in Wolfsburg with the mission to plan a new VW assembly plant
codenamed "B" in the Spanish Basque country. The plant was to be
die embodiment of the principles Lopez had followed at GM. Autos
were to be slotted together like a child's Lego set from prepackaged
modules rather than from hundreds of small components. The
workers engaged to slot these modules together would be supplied
from outside contractors at lower rates than those normally paid to
employees on an assembly line. The concept was remarkably similar to GM's aborted Project X.
Three months later, GM began a legal action against Lopez
alleging industrial espionage, its Opel unit in Germany claiming
that Lopez and the members of his team who had defected with
him had carried away crates of secret information that was copied
and entered into VW computers. So serious were the charges that
the German government began its own investigation, which led,
three years later, on December 13, 1996, to Lopez being formally
indicted. Charged along with him for industrial espionage, in a fiftyeight-page indictment, were colleagues Jorge Alvarez, Jose Manuel
Gutierrez, and Rosario Piazza.
When a top executive moves over to his competitor, having
someone sort of watch while he packs his office is net being
too pushy.
This was in many ways a classic case. An executive at one company moves to another company. He takes not only his know-how
and experience but also documents. Hard to protect against under
any circumstance, how do you keep your purchasing head in the
I dark about what's going on? One way for GM to have protected
itself was to compartmentalize information, to keep Lopez from
j1 knowing everything there was to know about what was going on.

224

THE

INTELLIGENCE

EDGE

The price for that, of course, would have been a secure, but paralyzed, GM. Youjust can't do that.
Lopez's crime was made easier, and much more damaging
than it would have been twenty years ago, by the invention of the
floppy disk. Assume that you had taken a filing cabinet full of documents. No matter how important they might be, a filing cabinet
could not possibly contain the totality of Project X, in all of its complexity and subtlety. But digitized data... well, a briefcase could easily contain the keys to the kingdom.

S E C U R I T Y A N D THE DIGITAL AGE

The Lopez case represents a classic problem of espionage by

betrayal, for which classic solutions are available. These problems
must be handled by the standard security procedures of a company.
But companies today have a new emerging set of security problems
that have not been anticipated and for which classic solutions will
not do.
A friend of ours in South Africa runs a merchant banking
house. He's had his ups and downs but, on the whole, has done
quite well. About a year ago, the South African government agency
that deals with securities firms investigated a merger of which he
was part. As frequently happens, stories of the investigation leaked
into the press and his name was tangentially mentioned. The newspaper in which it was mentioned implied that in some way he was
under investigation. That wasn't true, but it was, after all, one of
many stories about him, and while upset, he was only mildly concerned. Unfortunately, he had a problem that he hadn't anticipated
and that wouldn't have been a problem just a few years ago.
It happened that the newspaper that mentioned the investigation is one that is monitored by Reuters, the British news service
that frequently runs stories from newspapers around the world on
its international service. These newspapers are not necessarily the
most influential or important. Theyjust happen to be the ones with
which Reuters has agreements. Nor is it run on the standard wire
that most global media pick up, but on a specialized wire that does

Taking Cover

225

not have very deep penetration. So it was a pity, from our friend's
point of view, but no major cause for concern. It would all be forgotten shortly.
Unfortunately, this was not the case. This particular wire was
archived on several major on-line databases, including Lexis-Nexis
and FT Profile. This meant that every time someone put our
friend's name into one of these databases, this story, with the implication of wrongdoing, would come up. Since it was a minor story,
and basically inaccurate as well, there would be no follow-up. But
for as long as on-line services carried Reuterswhich is to say for at
least our friend's lifetimethere would be a story in there, along
with all his successes and awards, clearly implying that our friend
may well be a bit of a scoundrel.
Ever since Gutenberg laid the groundwork for mass media,
one solace has been paper's short legs and memory. Not only would
a newspaper story in South Africa not be known outside the circulation area of that newspaper, but would usually not even be
remembered in South Africa. And if it was vaguely remembered by
someone, who could possibly find it, even a few months after publication? There was a comforting element of forgetfulness about the
media. You could live things down.
This was no longer the case. Our friend came to the United
States trying to arrange financing for a new project. He was
approaching some major American banks where he wasn't well
known. He had good introductions, so he was being taken seriously,
but naturally, his hosts wanted to know a litde more about him, so
they checked his name in some, on-line services. The last story on
himit was the most recent and therefore the first to come up
was the story about the investigation. Plus, it turned out that
another British newspaper, picking up on the Reuters story, had
mentioned his name in another obscure, but archived, publication.
It appeared to his hosts, not knowing any better, that he had
recendy run into some serious regulatory problems back home that
hadn't yet been cleared up. Fortunately for him, one of the bankers
he was meeting with had the courtesy to mention the problem to
him, allowing him to not only explain it but to call his office and
arrange for his attorneys to gather some exculpatory material. But

226

THE

INTELLIGENCE

Taking Cover

EDGE

the fact remained, except for the opportunity to explain himself,

the new public record of on-line services would forever carry that
information. Every time someone checked him out, those stories
would be there, on-line, coming up every time his name or firm's
name was put into the query field.
In effect, this was a case of unintended disinformationelectronic domain style. Absolutely no one had planned to harm his
reputation, but there it was. Information that's emitted no longer
dissipates. It congeals. That means that our friend could not treat
that story indifferently. It could and did come back to haunt him.
The real division of the media today, from an intelligence point of
view, is the archived and the unarchived media. Information that is
released and archived becomes part of the permanent record.
Information that is unarchived does not. In the electronic archives,
it is extremely important to note, major publications like the London Times and minor publications like the Baton Rouge Advocate are
both available and have equal weight. If they carry someone's name,
they will come up not in order of prestige but with last publication
first. And we have noted that, psychologically, these electronic
archives are great democratizerswe read press releases with the
same care that we read New York Times columnists.
Now, if something accidental could cause harm, imagine what
deliberate misinformation, targeted at archived media, could do.
Imagine what unintended releases of information from your own
company can do. Today, the first line of counterintelligence is not
the secretary stealing documents from the filing cabinet to give to
her KGB lover, but enemies deliberately releasing information into
strategic, archived publications, there to haunt you forever. What is
said about your company and what your company says about itself
today has taken on massive significance. Where previously new
reports would have to reach a level of notoriety rarely achieved in
order not to be forgotten, today it has become impossible to live
anything down.
Business must take control of this electronic environment. In
the case of our South African friend, our advice was to issue a press
release. That sounds fairly trivial but it isn't. Today, press release
organizations, like Business Wire and PR Wire, exist, and for a few

I
I
I

227

hundred dollars they'll process and distribute press releases. Almost

all on-line databases archive these releases, right alongside major
newspapers. Many browsers aren't even aware that Business Wire is
a paid press release servicea vanity press, if you will. By issuing a
press release via Business Wire, our friend guaranteed that he
would control the first story that came up. He would also be in a
position to counter the impression left.
This seems an odd thing to discuss in the context of counterintelligence, but everything that we have been saying in this book
leads to the conclusion that the emerging information battleground is the electronic domain, where the very laws of information are being rewritten. This is where attacks will be made and this
is where they will have to be defended against. And ultimately, this
is where leaks will occur.
Consider this tragic tale: A computer scientist had been working on a breakthrough piece of software for several years, and had
been coming close to success. His company was getting ready to go
into high gear on his work. He attended a conference in Cologne,
West Germany, where he was on a panel as a commentator. He was
among friends; it was a small gathering; he was far from home; and
he was relaxed. He spoke a bit too openly about his workbut
then, really, who would know?
The conference was being transcribed by a company that transcribed conferences and sold the transcriptions to an American
technical publisher who then sold the transcriptions at absurdly
high prices. As part of its strategy, it also published some technical
magazines and frequently ran teasers from conferences. Fairly casually, the transcription editor selected an interchange involving the
computer scientist in the monthly magazine, which also was
archived on Lexis-Nexis. In short order, whether keying in the type
of software, the scientist's name, or his company, the most recent
thing to come up were his commentshighly revealing to the competition. The competition, also working in the same direction,
rapidly shifted plans and sped up its own work, beating the scienlist's company into the marketplace by months. Disaster.
The first step in contemporary counterintelligence is a dramark shift in the psychology of employees concerning information.

228

THE

INTELLIGENCE

EDGE

This is infinitely more important than nifty badges dangling from

the neck on chains or keypads on the door. An awareness that nothing is casual any longer. The recognition that emissions of information no longer dissipate as they once did has taken a long time to
penetrate corporate culture. The emphasis there has been on traditional security.
Making sure that only authorized personnel enter the
premises.
Making certain that critically important documents are seen
only by those authorized to see them.
Making certain that copying machines and secret spaces are
kept separate,
Sweeping the building for bugs.
Checking the briefcases of people entering or exiting buildings or areas.
Maintaining the integrity of computer systems.
All of these remain extremely useful. But the basic modern
threat does not come, in our estimation, from hackers breaking
into your network, While that can happen, it can also be guarded
against. The basic threat to the security of your information comes
from uncontrolled and unguarded emissions from your own
company.
This has always been a problem. As the Lopez story points out,
one of the major vectors of illicit emissions is people motivated to
betray. But the real threat today comes from the loyal employee,
doing his best at his job. The weak point is the casual comment to
a reporter, an indiscreet response at a professional conference, a
professional paper not clearly thought through.
Nothing dissipates. You must learn to live with every
indiscretion as long as electrons remain electrons.
One of the interesting phenomena we have noted has been
the proliferation of local business papers. Even relatively small

Taking Cover

229

towns have spawned business journals and their imitators. These

newspapers carry tremendously detailed, friendly stories on local
businesses. Executives, speaking to a friendly audience in the editor
and reporter, tend to let their hair down and speak frankly to their
community. These local journals are gold mines for us, carrying
detailed information on companies and people we could never
find out about from the national mediaand they are on-line
alongside the Wall Street Journal. The speed with which information congeals on-line creates a real-time threat to your company.
Almost before you know it, an indiscreet remark is embedded in
databases.

BASIC S E C U R I T Y IN THE DIGITAL A G E

In the olden daysabout ten years agomost companies maintained safes in which vital and sensitive documents were stored.
Today, the most valuable information in a company exists in endless, unlocked litde cubicles called hard drives. What is amazing
about our hard drives is not only how much utterly invaluable information these drives contain, but how little most managements
know of what is contained on them or where. The only thing more
amazing is how utterly insecure these drives are.
Any effective security and counterintelligence system must
begin with the hard drive. First, companies must, once and for all,
develop systems for finding out what employees are keeping on
their drives. This is not a matter of making certain that employees
are not playing Doom on company time, but of developing an effective inventory of a company's knowledge base, increasingly present
on the hard drive more than anywhere else. Tools are now becoming available to allow managers to access hard drives. Windows 95,
for example, provides connectivity between the hard drives. One
result is that, for the first time since personal computers swept into
offices, it has become possible to take inventory. It also becomes
easier to tap into other people's hard drives and make off with what
they have.
The hard drive isn't a stockroom. You can't take physical

230

THE

NTELUIGENCE

EDGE

inventory and be certain that nothing has been taken because it is

easy to copy information without the copying being noted: You can
put files on a floppy, you can send it via e-mail, or yon can just pick
up the hard drive, take it home, and copy the material to another
hard driveduplicating valuable information infinitely. Indeed,
the hard drive's vulnerability rests in the fact that it can simply be
stolenphysically taken, with or without the rest of the computer.
A hard drive is extraordinarily small for the value it might contain.
So we are in a situation in which a majority of a company's information rests on what is probably the least secure part of the computing system.
There are three basic solutions to hard-drive security:
1. Passwords
2. Encryption
3. Scrubbing the disk
None of these solutions, by itself, will suffice. Passwords protecting
access to either files or to the computer system itself certainly will
stop the untrained and may slow down the experts, but they can be
broken. The passwords systems protecting major word-processing
software files can and have been broken. The same is true for
encryption. Encryption can always be broken, given enough time
and enough money Encryption can be so complex that it would
take the National Security Agency twenty thousand man-hours to
break the codeor it can be so easy that a fourteen-year-old hacker
can get past it in an hour. Encryption does not absolutely protect
anything. What it does is increase the amount of time and the
resources required to break into the file. Obviously, this means that
the more valuable the prize inside, the more likely that someone
competitor or law enforcement agencywill go to the trouble of
breaking the encryption seal.
In the end, the best protection for files is removing them physically from the hard drive and storing them in a sale place that no
one knows about. While it certainly defeats one of the reasons for
having the hard driveconvenienceit does secure vital information. But even deleting a file is not quile as easy as it sounds. The

Taking Cover

231

act of deleting a file does not erase it. The data is still there and can
still be read, sometimes by simple utilities designed to undelete
and sometimes by more sophisticated means. All deleting a file does
is tell the operating system that the space occupied by the file is
now available for use. Until the space is actually used, the file is
still there.
Consider this: While you are clicking away in Microsoft Word
or Corel Word Perfect, backups are taking place. Every fifteen minutes or so, whatever you are working on is backed up. Fifteen minutes later, the backup file is deleted and another one is created.
Every time you print the file, a print file is created on your hard
drive. In the course of writing a fifteen-page proposal, literally hundreds of copies of that file have been created and deleted. That
means that for weeks or even months after you have mailed out the
proposal, deleted it from your drive, and forgotten about it, enough
copies of that proposal remain on your drive that it can be reconstructed with not very sophisticated means. The larger your hard
drive, the more free space it has, the more likely past secrets and
indiscretions are still sitting on your hard driveeverything from
e-mail to secret sketches of your newest project.
First off, tie down your central processing units (CPUs) and
put strong, serious locks on them so that your computers can't take
a walk or have a hard drive lifted out of them. Second, put in place
a systematic scrubbing approach. Using software like "defraggers,"
which consolidate files on hard drives, it is possible to overwrite all
unused sectors, thereby eliminating old files. But bear in mind that
recent thinking is that a single overwrite will still allow sophisticated
systems to "read" your files. Plan to overwrite several times. This is
a bit time consuming, but it will mean that even a stolen computer
won't yield your most valuable or dangerous secrets.
The network is the next problem. Again, the very way in which
it has made life easier also makes information more vulnerable. An
unscrupulous employee, for example, can e-mail your secret business plan to your competitor, going into the server drive where it is
stored and shipping it out. Some simple rules are essential:
Access to server drives is a great convenience. Creating multiple real or virtual drives and controlling and limiting access

232

THE

INTELLIGENCE

EDGE

to them is vital. Everyone should not have access to everything.

Password protection and encryption will slow down most
thieves. They will also provide you with hints of hacking. Logs
of failed passworded log-ins are extremely valuable. If someone tried to get into your account or a file twenty-four times
yesterday, it probably wasn't an accident. Protect.
Firewalls between your Internet site and the rest of the network are essential. The best firewall: no connection at all.
Use a separate computer off the network to serve the Internet. Seal the rest of the system off from the world, except of
course for e-mail.
Set rigid rules on not downloading software from the outside
to any computer on the net without using virus-scanning
software. Enforcing this rule by monitoring incoming and
outgoing files by size also has the advantage of warning you
when someone is trying to ship out large amounts of information.
Data transfers of all kinds must be monitored. Routine e-mail
is not the problemor at least not the major problem. The major
problem occurs when a multimegabyte document is shipped out.
Make it clear that the export of files larger than a certain size will
not be supported by your e-mail system and will require technical
support from the system manager. This will mean that whenever a
really humongous file is going to be shipped, someone else is
alerted and can examine it.
But what is going to kill you are the floppiesthose miserable
3.5-inch disks that can contain an entire book. And this is the heart
of the matterthere is never a foolproof security or counterintelligence system. Neither the Borgias nor Microsoft can defeat human
cupidity. What can be done is increase the time and cost of these
efforts. The goal can never be to make industrial espionage impossible. What it should be is to increase costs and risks beyond the
potential value of the information. That can be done, albeit with
cost to your business as well.

Taking Cover

233

CONCLUSION

In our view, businesses are threatened far less by the theft of their
information than by their own people giving that information away.
A casual remark in the hands of an analyst with context and nuance
is more dangerous to your business plan than fifty former Mossad
agents. And no amount of legal protectionnot copyrights,
patents, or confidentiality agreementswill protect you from the
careful collector and the meticulous listener.
This is the shift that is taking place in business intelligence.
The threat does not come from espionage but from intelligence;
not from breaking and entering as from leaking and gossiping; not
from unscrupulous former employees as much as from loyal current employees. And it poses a fundamental challengeand opportunityto modern business. The challenge is to protect your own
material and the opportunity is to see what your competitor
is doing.
The problem here is that the solutionstrict controls on the
movement of information within a companyposes a tremendous
price for business. Inefficiency grows out of secrecy, and inefficiency can destroy a company. The problem facing business today is
securing information without capitulating to secrecy. It is a far
graver challenge than the KGB ever posed.
The solution is twofold. First, it is cultural, recognizing that
loose lips do indeed sink ships. Imposing a culture is difficult, but
less difficult than imposing secrecy. Second, the solution is technological. Emerging, but not yet present, technologies will allow us to
tag information and know who has seen it and what they have done
with it. As information grows more precious, tracking its movements will become the solution to a problem that cannot help
but grow.

And Now a Word from Our Lawyer

AND NOW A WORD

FROM OUR LAWYER
Intelligence and the Law

Well, it's that time againtime to visit with the lawyer. Always an
interesting experience. Notice how he starts; the very first line: " 'Is
it legal?' That's the basic question people have about law and corporate intelligence." How do we convince him that this is most
emphatically not the basic question? The basic question is how can
I find out what I need to know and riot get into trouble? There's a
world of difference between the two questions.
The world of intelligence is a world of ambiguity and risk
you are never quite sure what the rules are but you've got to act anyway. In the world of law, you never act until you are sure of your
ground. But how can anyone be sure of his ground on a project
involving Somali and Chinese oil? What groundwe're in a swamp!
The lawyer's solution: Wait while he researches the legal aspects.
Our answer: Do it and let the lawyers sort it out later. Visits from
legal counsel are, in our world, exciting moments indeed.
We don't have contempt for the law. What we have is a real
problem! At any moment the law can shut anything down, and we
can't afford to be shut down. In a world where there are countries
in which asking about rice production can get you put in jail, we
need to find a way to do what we have to do without going away in
handcuffs. So the legal question is important but, from our point of
view, ultimately subsidiary. It's something that you need to ask to be
a responsible citizen, even if you get a long, convoluted answer that
basically says: "Maybe, but don't hold me to that."
234

235

Notice the little disclaimer in the third paragraph: "Without

here providing legal advice, we can offer some useful guidance to
the reader first by pursuing some clarifying questions." Is there
someone out there who can tell us what that's supposed to mean?
We guess it means that he doesn't want to be held to anything
he says.
But we're in the real world. So while our accountant has been
sweating bullets by our side through this whole thing, now our
lawyer comes in to get the last wordand leave it to us to figure out
what that word is.
We can't wait to see the bill.

THE L E G A L CONTEXT
BY JOHN S. B A K E R , JR.

"Is it legal?" That's the basic question people have about law and
corporate intelligence. The question can be asked in a way that
implies criminality: "Isn't corporate espionage a. crime?" However the
question js phrased, though, a competent, cautious lawyer typically
might answer: "Each situation is different, depending on the facts.
In many areas, the law is unclear. You should consult an attorney
with your specific questions."
There you have it; an answer that is perfectly accurate...and
completely useless! Worse than useless, however, is the answer of
the less-than-competent lawyer who, like many accountants, simply
says: 'Just don't do it!" That way, of course, the lawyer cannot be
blamed if anything goes wrong.
Without here providing legal advice, we can offer some useful
guidance to the reader first by pursuing some clarifying questions:
(1) Is intelligence gathering the same as criminal espionage?
(2) Even if not criminal, is intelligence gathering a legitimate corporate activity? (3) How does one recognize a criminal or illegal
corporate practice? Intertwined with these questions is another
one: Are we talking about American law, the law of some other
country, or both?
The terms intelligence gathering and espionage suggest that the

236

THE

INTELLIGENCE

EDGE

main legal issues involve criminal law, when actually the most significant issues involve intellectual property and trade secret law.
The United States, in particular, has focused on protecting innovation and preventing what is deemed to be unfair competition.
Among its many firsts, the U.S. Constitution authorizes Congress to
grant exclusive rights to authors and inventors. Congress has long
given patent and copyright protection for intellectual property and
more recendy has protected trade secrets that were already generally protected under state laws.
The Economic Espionage Act of 1996 criminalizes the theft of
trade secrets. What makes either intelligence gathering or espionage criminal under this law is the misuse of trade secrets. If the
crime is an act committed on behalf of a foreign government, it is
labeled economic espionage. Otherwise, the crime is theft of trade
secrets. If trade secrets are not involved, someone who employs
"espionage" techniques does not commit the crime of economic
espionageeven though the person is "spying" on economic activity. Thus, in the United States, "espionage" and intelligence gathering need not be criminal nor illegal.
Few countries, besides the United States, protect trade secrets
as such. Of course, a foreign government might charge theft or
embezzlement in a case involving trade secrets, as was filed in Germany against former General Motors executive Jose Ignacio Lopez
de Arriortua. Nevertheless, outside the United States, the secrets of
a private corporation do not have the same protection as they do
within the United Stales. Taking economic "secrets" from a foreign
government or a government-owned corporation will, however, very
likely be considered not merely theft but criminal espionage.
In less free countries, the scope of espionage is quite broad. A
country may treat as state secrets matters which in the United States
would be public information. Basic statistics about a country, such
as population and industrial output, are still technically illegal to
obtain in many areas of the world. In former Soviet bloc countries,
such laws may or may not be enforced. A practical way of gauging a
country's interpretation and enforcement of criminal espionage as
to basic economic data may he to assess the degree of freedom
afforded journalists and lawyers to practice their professions. A

And Now a Word from Our Lawyer

237

country such as China, which has had very few lawyers and whose
government greatly restricts and sometimes arrests journalists, is
not a hospitable place to conduct private intelligence gathering
at least on the ground.
Lawyers, as well as reporters, around the world routinely
engage in intelligence gathering. American trial lawyers call such
fact-finding or investigation "discovery," a process governed by rules
and enforced by courts. Any lawyer or reporter who does not adequately investigate before making allegations acts irresponsibly. For
lawyers and journalists, intelligence gathering must be considered
obligatory.
If intelligence gathering is permissible when done lawfully by
lawyers and journalists, should it be any less so when done by corporations? If there seems to be a difference, the explanation may lie
in the labeling rather than in any analysis. Legal and journalistic
research may seem more legitimate than corporate espionage
because lawyers and journalists, although acting in a private capacity, claim to advance the public interest. Is there really any difference, though, in what these professions in fact do and what
businesses do when it comes to intelligence gathering?
What is the difference, actually, between some forms of
research and espionage? Is it research or spying (or both) when an
insurance company investigator videotapes the physical activities of
a person who has made a claim for injuries allegedly resulting in
permanent disability? Is it research if directed by the lawyer but spying if done by the corporation? Are the labels at all meaningful?
Cannot the intelligence gathering done by a corporation benefit
the public, just as it can when done by a lawyer or journalist?
Intelligence gathering, practiced legally, is part of the competitive enterprise. A person or corporation that is the object of intelligence gathering may call it corporate espionage. But what if that
corporation takes defensive security measures? Aren't both corporations then gathering intelligence on each other? Are they both
engaged in corporate espionage? Is one or both of them engaged
merely in market or competitive research? The answer given at the
beginning of this chapter by "the competent, cautious lawyer"
that is, "it all depends"has this virtue: It resists the tyranny of

238

THE

INTELLIGENCE

EDGE

labels. Some forms of intelligence gathering, such as trial-related

discovery, are not only legal but legally enforceable. At the other
end of the spectrum, some forms of intelligence gathering, such as
theft, are not only illegal but criminal. The legality and criminality
of most intelligence gathering, however, does "depend"... namely
on the rights to the information, the means of obtaining it, and the
use to which it is put.
Illegality and criminality are often inaccurately equated. Not
everything illegal is criminal. The fact that an act is not criminal
does not mean it is legal. Many "wrongs," such as personal injuries,
are not legal or right, but they may not be criminal. Personal
injuries are most often "righted" by civil remedies, usually monetary
damage awards. It is, however, more difficult to distinguish the
criminal from the merely wrongful or illegaland even from the
innocentwith business-related offenses than it is in personal
injury cases.

LEGAL RISKS I N PASSIVE I N T E L L I G E N C E

Passive intelligence gathering involves the lowest level of legal risk.

Legal issues revolve around rights in information. Legal issues can
arise at each point of the passive research process: reading, copying,
and manipulating. Most commonly they arise in the context of
copying materials.
Basic Copyright
Just think about how many people have ever made a photocopy of
a copyrighted document. It need not have been a secret documenta newspaper article qualifies. Such copying is prohibited by
U.S. federal law and, by treaty, must be prohibited by signatory
countries. Even a single instance of copying copyrighted material,
unless it falls within the "fair use" exception (determined by four
factors, including the amount of material and the impact on the
economic benefit to which the copyright holder is entitled), is illegal. Under U.S. law, it also becomes criminal if done "willfully and

And Now a Word from Our Lawyer

239

for purposes of commercial advantage or private financial gain" (17

USC Sec. 506 fa]).
Even civil enforcement, which is relatively rare, of noncriminal
violators has been most likely when financial gain is involved.
Kinko's, a nationwide and now international copy service, was sued
by several publishers for copyright infringement for copying and
printing, withoiit permission or payment, selections from copyrighted books that were collected into anthologies and sold to students for use in college courses. In 1991, a federal court rejected
Kinko's claim that the copying constituted "fair use" and found that
Kinko's had committed copyright infringement. The court also
stated: "Kinko's has failed to prove it was an innocent infringer by
failing to show its good faith. Kinko's should have known that it was
infringing plaintiffs' copyrights."
The practice at issue in the Kinko's case involved publishing
rather than intelligence gathering. Yet the case has had an impact
on all commercial photocopying practices. Since losing the lawsuit,
Kinko's will not copy a single newspaper article or more than one
page of a book (deemed to be within "fair use") without a written
copyright permission. The civil lawsuit it lost does not dictate that
policy. The copyright statute itself imposes that obligation. As a
result of the lawsuit, obviously. Kinko's has become more sensitive
to its statutory obligations. As a commercial "gatekeeper," Kinko's
has become an effective enforcer of federal copyright law against
others who use its services. If Kinko's did not now do so, it would
not only risk civil suits but would become a more likely target for a
criminal prosecution. Its personnel can no longer claim ignorance
that its prior business practices-violated the law. The litigation eliminated any arguable legal ambiguity as to certain business practices
and prompted greater caution as to other practices.
Electronic Copies
Transforming data and other information into intelligence requires
copying. As long as manual, typed, and photocopying were the only
widely available ways of copying information, intelligence gathering
outside government was limited. As in so many matters, computers

240

THE

INTELLIGENCE

EDGE

have changed the landscape. The same computers both make private intelligence gathering cost effective and also make illegal copying much more cosily to the owners of information. One does not
imply the other, however. Intelligence gathering does not require
illegal copying.
Normally, mere reading of (at least, nonclassified) information raises no legal issues. Certainly this holds true when reading a
book or other printed publication. Unless you have stolen the publication, you need no legal permission to read and thereby obtain
information. Stating the obvious would not serve any point, but for
the fact that the same treatment almost did not apply to electronic
materials.
In December 1996, treaty negotiators reached two new international agreements on changes in copyright protection. Negotiators rejected attempts to make reading copyrighted material on a
computer screen a violation of copyright law. Few people would
think it should be. But there were those who would have made
anyoneincluding schoolchildren at their librarycopyright
infringers for even looking at the copyrighted material on a computer screen.
What makes research done on-line different from that drawn
from a book? With a book, the acts of publishing, reading, and
copying are clearly distinguishable. The publishing of a book is
complete before the ordinary reader has the opportunity to read it.
If the reader copies part or all of the book by hand or using a copy
machine, he clearly must make the decision to do soregardless of
whether the act of copying is or is not permissible.
In electronic publishing, the lines separating publishing, reading, and copying are blurred. If you access a commercial on-line service such as LEXIS-NEXIS, you pay a fee that may include only
reading, both reading and copying ("downloading"), or separate
fees for each. You may be paying to read materials that you might
otherwise read without cost in hard copy, but it may be difficult or
time-consuming to get the information. So you pay for instant
access. Paying the service provider effectively satisfies whatever
financial obligations are due the copyright holder (or other owner
if the material cannot be copyrighted). The provider pays the copyright holder.

And Now a Word from Our Lawyer

241

The openness of the Internet invites copyright infringement.

Internet-service providers, through which users access the Internet,
have no affirmative obligation to screen out illegal copies posted on
the Internet. As a result, in the absence of the 1996 international
treaty, users of the Internet arguably are routinely violating copyright. Browsing the world wide web necessarily involves producing
a temporary copy of the material to be read by the user. Thus electronic publishing, reading, and copying are not neatly distinguishable. While the treaty exempts the temporary copying inherent in
basic electronic viewing from infringement, it leaves vague the distinction between a temporary and a permanent copy.
Software and Databases
Electronic copying, as already indicated, does not basically differ
from other kinds of copying as a matter of copyright law. Because
of the amounts of revenue lost to software producers, what differs
is the enforcement of the law. Software piracy violates both the
copyright law and the trade secrets law, as discussed below.
Any business using computers has to be concerned about its
employees using illegally copied software for any business purpose.
It does not matter that the use of illegal software was unauthorized,
unless the copying or use was contrary to clearly established
employer policy. If the business is not selling illegal copies, it might
seem it has nothing to worry about. Not true. Unlike photocopying,
relatively small-scale illegal software copying does receive vigorous
enforcement due to the intelligence gathering of (he Software Publishers Association (SPA). To protect themselves, businesses need to
implement their own internal security to prevent being the object
of enforcement.
The rarity of enforcement actions for illegal photocopying as
well as the legality of personal-use-only video-copying (e.g., recording a televised movie on a home VCR) seem to have misled many
into casually making illegal copies of software. Like the music industry trade associations that have engaged in enforcement for years,
the SPA directs an aggressive effort to identify, punish, and deter
software copyright infringers. In doing so, the SPA effectively
employs passive, semiactive, and active intelligence gathering. As of

242

THE

INTELLIGENCE

EDGE

1995, the SPA's toll-free telephone line has been receiving a hundred tips per week, generally from disgruntled employees, identifying use of illegally copied software. The SPA verifies reports by
checking whether the targeted company has registered with the
software publisher. If the company has not, the SPA will seek a voluntary audit in which the company agrees to allow the SPA to
inspect its computers for pirated copies. In doing its audit, the SPA
uses special software that can detect whether any software has been
deleted.
Why would any company allow the SPA such access? It's quite
simple. As a result of amendments to the copyright act in 1992, the
burden of proof rests on the software user. The user must, by
receipts and/or other records, prove its right to use the software. If
the company does not agree to a voluntary audit, the SPA can seek
a court order for an involuntary audit. When that occurs, the SPA
arrives with federal marshals to execute a search warrant to inspect
the computers. Either way, if the company is using illegally copied
software, it will pay fines. If the company resists, officers and/or
managers may suffer criminal prosecution.
Intelligence gathering involves the use and creation of databases. Copyright protection does not cover facts, only the manner
in which they are compiled. The user would prefer that databases
not be copyrighted, and courts have traditionally restricted severely
the scope of database copyright. That allows users of data more
freedom to create new databases. When the user creates a new database that has commercial value, however, the user-turned-creator
may now have second thoughts about copyright coverage.
The 1996 Geneva conference on copyright failed to expand
copyright protection to databases. Compilers of databases would
like the same protection for their work product as is currently provided only for creative materials. Like software producers, database
owners can and do charge for access. Without copyright protection
for a database, however, the user is not barred from copying,
manipulating, and/or redistributing the dataunless restricted by
agreement.
Database creators may be able to restrict their use by contract,
but the protection may still be lacking for some. Market conditions

And Now a Word from Our Lawyer

243

and the database owner's marketing strategy will determine its ability contractually to impose the conditions. The distribution requirements of certain databases preclude charging fees, much less
imposing other conditions. Obviously, for example, the New York
Stock Exchange must continue to distribute its information to millions without charge and therefore cannot contractually impose
restrictions on those who read its quotes in newspapers. With copyright protection it could continue to distribute to the same audience without charge, but the law would impose a restriction on
copying, manipulation, and redistribution without permission of
the copyright holder, which presumably would involve a charge.
Theoretically, copyright protection and the ability to charge
should stimulate the creation of more databases, thereby increasing
the sources from which to gather intelligence. Increased supply and
competition supposedly should more than compensate for the fact
that some currently available free sources of information would
carry a charge. On the other hand, with or without a charge, databases would become less generally accessible. Although lawfully
obtained, databases, like stock exchange statistics, would no longer
automatically be available for incorporation into newly created
databases. The ease and cost of accessing and creating databases is
central to passive intelligence gathering.

LEGAL RISKS OF THE S E M I A C T I V E

A N D A C T I V E MODES
^

Semiactive intelligence gathering, like telephone market research,

does not involve any particular legal risks in free countries as long
as fraud or bribery is not involved. Active intelligence gathering can
simulate and may supplement or substitute for law enforcement.
Naturally, it carries the greatest legal risks. As exemplified by the
theft of trade secrets, the legal issues involve overlapping questions
related to rights in information, the means of obtaining it, and
its use.
The problem of trade secrets demonstrates that at least for
defensive purposes, corporations with valuable proprietary infor-

244

THE

INTELLIGENCE

EDGE

mation must undertake security measures that, in themselves,

amount to active intelligence gathering. Moreover, if certain corporations examine their policies related to security they will realize
they are pursuing inconsistent goals. On the one hand, computerrelated as well as other corporations are resisting federal proposals
to limit encryption. Federal law enforcement and intelligence
agencies want access to encryption software in order to subject
electronic communication to court-ordered surveillance. Simultaneously, the same industry has pushed for new federal criminal laws
on trade secrets and computer crimes. The obvious argument by
federal agencies is that it has been given more responsibility to solve
computer-related crimes without commensurate ability to do so.
The solution for corporations may be to rely more on prevention
than on federal criminal enforcement.
Requesting Information
The semiactive modefor example, making telephone calls or
conducting face-to-face interviews for informationdoes not itself
ordinarily involve any particular legal risks, at least in tree countries. In China, on the other hand, asking routine questions about
business activity may be deemed a subversive act and therefore
criminal activity. A Western businessman attempting to track down
the source of piracy of his product may find himself accused of espionage simply for asking. Even within liberal democracies, laws do
protect the secrecy of information stamped "classified" or "state
secrets." No country is more open than the United States and yet
the shear size of our federal government means that more information is classified than the total amount of information many
countries have. As a result of the end of the Cold War, as well as the
simple passage of time, which eliminates all but die historical value
of most information, die reasons for classifying much information
have been lost. Federal agencies are currently undergoing a complete review to declassify information that no longer needs to be
classified, but the process is expected to take a number of years.
In the United States, the presumption favors disclosure of
information. Government information may be classified, but it can-

And Now a Word from Our Lawyer

245

not be copyrighted. Once disclosed it becomes freely available.

Moreover, any citizen can request nonpublished information from
almost any federal agency following the relatively simple procedures provided by the Freedom of Information Act (FOIA). Such
requests produce a great deal of information that agencies would
not have volunteered to release, if only for reasons of inertia.
What about paying for information? What's the difference
between paying for information and bribery? Paying another's
employee in order to gain business from the employer may constitute commercial bribery in the United States. If the transaction
includes the transfer of proprietary information, theft of trade
secrets may also be involved. The original meaning of bribery,
which applied only to payments made to government officials, has
expanded in the United States to commercial situations. Definitions of commercial bribery are quite broad, but fortunately the
crime does not generally get applied to exchanges of gifts and
favors among businesspeople, which, according to prevailing custom, are normal and acceptable. Otherwise, under the wording of
some American commercial bribery statutes, a salesperson who, as
part of a sales call, buys lunch for a corporation's purchasing manager, literally could be guilty of commercial bribery.
American corporations conducting business outside the
United States have greater cause for concern about "bribery," in the
traditional sense of payments to government officials. In many
countriessuch as Mexico, where the governments have substantial control over the economy"customary" payments are
expected. The antibribery provisions of the Foreign Corrupt Practices Act, however, make it a crime for U.S. businesses or their
agents direcdy or indirectly to pay or to give anything of value to a
foreign government official, any person acting on behalf of a foreign government, any foreign political party, or candidate for office
in a foreign country. The statute does not prohibit payments to others unless some or all of the payment is going to a covered person
or entity. The prohibition applies to payments for "influencing any
act or decision of such foreign official in his official capacity." Those
words standing alone, however, include not only clear cases of
bribery but other kinds of "customary payments."

246

THE

N T E

G E N C E

EDGE

Recognizing problems with the statute, in 1988, Congress

amended it to clarify that payments "for routine governmental
action" are not prohibited. These include "any facilitating or expediting payment" for such services as obtaining work permits,
processing visas and work orders, providing police and other government services, as well as phone service and other utilities. The
exception for such "facilitating" payments does not solve the basic
competitive disadvantage American companies face against companies willing to pay genuine bribes. The exception, however, does
distinguish and permit customary "grease" payments.
Tape Recording and Wiretapping

Among the familiar techniques common to intelligence gathering

and other purposes is the recording of telephone conversations.
Such recordings may be routine or for a special purpose. As the
Nixon tapes prove, their impact can be for posterity.
The legal treatment of recording telephone conversations
generally turns on whether one party consents to the recording and
whether it fits within the expanded notion of a "wiretap." Government wiretapping is easiest to pin down. Both the U.S. Supreme
Court, under the Fourth Amendment of the Constitution, and the
European Court of Human Rights, under Article 8 of the European
Convention on Human Rights, have ruled that government wiretaps must proceed according to written regulations that restrict
their operation and require advance approval by a judicial officer.
In the United States, it is a crime for anyone other than a prop
erly authorized government official to engage in wiretapping. Federal law does allow (although some states make it a crime) for one
party to record the other's conversation without the second party's
knowledge. Federal constitutional law treats consensual and nonconsensual recording differently; the distinction turns on the
notion of "reasonable expectation of privacy." When neither party
consents to the recording, they both have a reasonable expectation
of privacy. At the same time, in speaking with others, we each run
the risk that the other will record the conversation and therefore
have no unilateral expectation of privacy.

And Now a Word from Our Lawyer

247

Federal statutory law, however, has been expanded beyond

"reasonable expectations of privacy" to prohibit recording a cellular or portable telephone conversation. Even though courts have
ruled that persons do not have "reasonable expectation of privacy"
in such over-the-air conversations, the federal wiretap law prohibits
intercepting a cellular telephone conversation. Had it not been for
the publication of a recorded cellular conversation involving House
Speaker Newt Gingrich, relatively few people would be aware that
such nonconsensual recordings violate federal law.
In the United Kingdom, the law on electronic eavesdropping
has been the recent subject of debate. In January 1997, the government failed in an attempt to allow wiretaps without judicial
approval. Also, reaction to interception and publication of embarrassing cellular telephone conversations by Prince Charles and
Lady Diana has prompted calls for greater restriction on various
types of surveillance by nongovernment snoopers. In the meantime, private surveillance in the United Kingdom may involve relatively minor civil or criminal violations that appear to go largely
unenforced. All kinds of private surveillance equipment are widely
available in the United Kingdom.
According to a major supplier of private security and intelligence technology, Quark Research Group, most countries in the
world (other than the United States) either do not have or do not
enforce restrictions against use of private surveillance equipment.
The Persian Gulf countries, however, are an exception at least in
that such equipment is not generally obtainable there.
Taking and Securing Information

Breaking into private or government property and stealing are

crimes around the world. The case of the Watergate break-in (done
for the purpose of obtaining information) attests that in the United
States even an illegal entry directed by government officials is
a crime.
Increasingly, corporations have to be concerned about "breakins" searching for information from computers and employees. As
noted above, few countries other than the United States legally pro-

248

THE

NTEULIGENCE

EDGE

tect trade secrets. Whether inside or outside the United States,

however, the protection requires a corporation to defend any information it wishes to keep secret. Outside the United States, "stealing" trade secrets may not be stealing at all. Until recendy, even in
the United States, theft has involved the taking of tangible, movable
property. Such property can only be in one place at one time. When
taken, the owner or possessor has clearly lost something. With the
taking of intellectual property, nothing tangible need be taken.
Lost is the exclusiveness to the information. The loss is the decline
in market or potential market value. Those who take may not think
they are depriving the owner of" anything, even though the taker
obviously realizes a gain of something.
The treatment of trade secrets is changing. Until recently,
legal enforcement in the United States was under state law and
generally through a private civil action. The increasing importance
of intellectual property as a significant source of wealth has
increased the incidence and consequences of trade secret theft.
As a result, in 1996, the U.S. Congress passed and the president
signed the Economic Espionage Act, previously mentioned, which
makes trade secret theft a federal crime but does not provide a
private right of action for damages. The new statute will mean
that large-scale incidence of trade secret theft will likely result in
indictments.
The new federal statute presents some problems of definition.
Although it utilizes the definition from the Uniform Trade Secrets
Act, the fact that the act is criminal and federal raises the uncertainties about enforcement. The statute defines trade secrets as
follows:
all forms and types of financial, business, scientific, technical, economic, or engineering information, including
patterns, plans, compilations, program devices, formulas,
designs, prototypes, methods, techniques, processes, procedures, programs, or codes, whether tangible or
intangible, and whether or how stored, compiled, or
memorialized physically, electronically, graphically, photographically, or in writing if

And Now a Word from Our Lawyer

249

A. the owner thereof has taken reasonable measures to

keep such information secret; and
B. the information derives independent economic
value, actual or potential, from not being known to,
and not being readily ascertainable through proper
means by, the public
Under the definition, the information need not actually be secret.
It is only necessary for the owner to have taken "reasonable steps"
and that the information have economic value "from not being generally known or being readily ascertainable through proper means
by the public."
Every good employee doing interesting work will "gain valuable learning experience." If every employee in a corporation routinely must sign a nondisclosure agreement, which is good practice
at least in technology companies, does that mean that everything
learned on the job by every employee falls within the definition of
a "trade secret" for purposes of this criminal statute? No employer
would have previously thought to sue an employee for walking
away with the valuable training provided by the employer, no matter how ungrateful or disloyal the employer thought the former
employee. It would have been cosdy to do so and the suit would
have been deemed frivolous. Even with an employee nondisclosure
agreement, the former employer would rarely sue the new
employer for benefiting from what the employee learned with the
former employer unless it was clearly well-guarded proprietary
information.
The increasing ease with which information is moved has
inspired greater concern for security precautions to avoid having one's communication "overheard" or "secret" information
copied or stolen. Connecting computers to the Internet facilitates
illegal copying and other crimes, especially fraud. These crimes and
the general desire for secure communication have created a
demand for copy avoidance. This requires encryption, which in
turn has alarmed federal law enforcement that their own intelligence gathering will suffer as a result of perfect, private security.
Given die enormous consequences of electronic copying and en-

250

THE

INTELLIGENCE

EDGE

cryption, they have received considerable attention in terms of legal

enforcement.
Using Information Transnationally
Growing international commerce and electronic communication
have propelled the revival of" an international law of commerce.
This development updates the lex mercatoria, or "law merchant," that
was a medieval body of customary rules used by merchants to supplement the commercial law of states. The practice never died out
completely. As modern nation-states rose in power, however, such
customary law was not viewed as "real law," which was deemed to
emerge only from legislative bodies. Today, the creation of an international commercial law has been proceeding largely, but not
exclusively, by multilateral treaties. The World Trade Organization
and also recent treaties on intellectual property designed to facilitate commercial development of the Internet are both products of
and creators of such international commercial law. Other sources
of the growing international commercial law derive from assertion
of extraterritorial jurisdiction by one nation, namely the United
States, and by contract, which is a method the European Union may
use to protect data transfers.
The U.S. contribution to shaping this international law of
commerce has not only been by treaty. Whether or not consciously
intended for this, the extraterritorial assertion of jurisdiction over
financial crimes and drug-related offenses has brought tremendous
influence globally. Outside of extradition and the limited area of
war crimes and genocide, criminal law has not been the subject of
common or reciprocal international enforcement. Nevertheless,
through the federal crime of money-laundering, the United States
directly asserts broad jurisdiction over nonresident, non-U.S. citizens, thereby affecting the practices of financial institutions around
the world. Such influence will only increase because, since 1990,
the Financial Crimes Enforcement Network of the U.S. Treasury
Department has been using and building the ultimate intelligencegathering tool ("A hybrid between a data base and a focused surveillance tool, it combines qualities of many surveillance

And Now a Word from Our Lawyer

251

technologies." See Steven Bercu, "Toward Universal Surveillance in

an Information Age Economy: Can We Handle Treasury's New
Police Technology?," Jurimetrics Journal, Summer 1994, p. 5), which,
in cooperation with Interpoi, "will consolidate, analyze and disseminate data concerning financial crimes throughout the world." *
The American assertion of extraterritorial antitrust jurisdiction, even of a noncriminal nature, has been a source of conflict
with other countries. The European Union also claims extraterritorial jurisdiction over competition issues. Nevertheless, the competition laws of the European Union as well as Japan are neither as
strong nor privately enforceable as are American antitrust laws.
Both the United States and Great Britain are free-trade
nations. Nevertheless, the two countries have had serious conflicts
as a result of the assertion of American antitrust jurisdiction.
Antitrust conflicts between the United States and Great Britain have
largely, but not exclusively, concerned airline service. Both countries have been protecting the interest of their dominant airlines.
The two governments have had a bilateral agreement regulating
the amount of airline traffic between the two countries. The United
States has long pushed for an "open skies" agreement that would let
market forces determine the access of airlines to each other's markets. Since the late 1970s, when airfares were deregulated in the
United States, competition from American airlines and pressure
from the U.S. government have prodded gradual deregulation in
Europe.
The British airline Virgin Atlantic has from its beginning battled with British Airways. Virgin was formed after an earlier British
low-fare airline, Laker Airways,'went into bankruptcy. In the early
1980s, the U.S. Justice Department launched an antitrust investigation to determine whether British Airways and other competing airlines in the United States and abroad colluded to put Laker out of
business. Also, Laker's British liquidator in bankruptcy filed an
antitrust lawsuit in an American federal court.
As a result of diplomatic negotiations, the Justice Department
dropped its Laker investigation in 1984. That did not end the Amer*N. Brady., 1/12/93

252

THE

INTELLIGENCE

EDGE

ican lawsuit by the British liquidator, even though the litigation was
blocked in part by a British court. Also in 1984, in order to avoid an
antitrust lawsuit from the newly launched, privately owned Virgin
Adantic Airways, the then-government-owned British Airways asked
for and received a letter of assurance from the Justice Department
diat certain proposed fares did not pose a possible antitrust violation. Since being privatized in 1987, British Airways has been sued
by Virgin Adantic Airways in a New York federal court. The court
rejected various legal arguments by British Airways that it should
decline jurisdiction and send Virgin Atlantic to a court in the
United Kingdom. Certainly, the American airline industry has had
a significant impact on the British airline industry not merely due
to the market but due to enforcement of American antitrust law.
The European Union (EU) may soon be exerting a strong
influence on American practices regarding the use of data not
through criminal law but through contracts necessitated by the EU's
privacy rules. Countries in Europe have adopted "data-protection
laws" designed to restrict the disclosure of personal information. As
a result, several countries have begun restricting the export of data
to persons or entities in countries that do not have sufficient protection for personal data. The United States is not considered as having sufficient privacy protection for personal data. A recent study of
U.S. law prepared for the Commission of the European Communities recommends solving the problem of inadequate and inconsistent standards in the United States not by blocking the flow of data
to the United States but by strict liability on European companies
for adherence to the European standards.
In summary, whether you are doing business in your local community or overseas, intelligence is a necessity. Wherever you do
business, law is a reality. To the question "Is intelligence gathering
legal?," the answer is "Yes, depending..."depending on what you
are trying to find out, where you are looking for it, and how you
plan to get it. In general, passive techniques in the advanced industrial countries are going to be legal. Passive techniques in some less
developed countries may be illegal, but as long as what you do is
passive, you're probably going to be safe from retribution. Legal
problems develop as you become increasingly active. First, you face

And Now a Word from Our Lawyer

253

dangers because you must act rather than simply absorb. Second,
you come into direct conflict with others who may use the law as an
instrument to stop you. This is particularly true in the United States,
the most litigious nation on earth. The legality of intelligence gathering intersects with the likelihood of a lawsuit merely to deter you
from pursuing the truth. So, intelligence gathering is legal, but...
Consult your attorney.

CONCLUSION

OK, is everything clear now? We thought so.

Actually, there are several important things to take away from
this discusssion. First, there is not a separate area we might call business intelligence law. Since business intelligence touches on everything, all aspects of law impinge. The obvious areascopyright,
privacy, wiretapare there. But so are antitrust laws and virtually
any other area businesses are interested in.
Second, Anglo-American law takes itself seriously. Legal systems in other parts of the world do not. The fact that Albania has a
law on the books does not mean that Albania intends that law to be
enforced. Conversely, the fact that Albania doesn't have a law on
the books doesn't mean that law enforcement agents won't nail you
for doing something they don't like. The problem is, of course, that
a citizen of the United Kingdom or the United States is dealing with
a government at home that expects businesses to obey laws overseas, even if the natives know that the law is a joke. Even obeying the
laws of friendly, legally and culturally compatible countries like the
United Kingdom and the United States is difficult.
Third, the growing power of multinational organizations like
the European Union and the World Trade Organization place businesses in a difficult position, particularly in areas like intellectual
property law, where the rules are not only not yet clear but actually
contradictory.
Fourth, with the emergence of the electronic domain, any
comfortable assumptions you may have about the law should be discarded. The law of passive, electronic domain intelligence doesn't

254

THE

INTELLIGENCE

EDGE

exist. Unfortunately, you can't wait for the law to define itself in
order to do your business. Your work goes on regardless of ambiguity.
This is the ultimate problem faced in dealing with the law. The
fact that business intelligence techniques are at the cutting edge of
a technological change that the law has not yet caught up with; the
new global nature of business intelligence; the intersection of business intelligence with fundamental international disputes and fault
linesall mean that the law is not a particularly faithful guide.
No matter what you do, somewhere, somehow, it may be illegal. And yet you must gather information or die.
Where neither law nor custom is a guide, we urge common
sense.
Be aware of the ambiguous position you are in.
In collecting information, try to make sure that you really
need it.
In collecting information, try to make certain that you are
not stepping on the toes of people who are extremely powerful in countries where the law isn't taken seriously, and where
you have no choice but to visit.
Try, for as long as possible, to stay with passive techniques in
the electronic and paper domain.
Not a very strong position, but the best available at this point
in time.

APPENDIX

TOOLS FOR FINDING PEOPLE

There arc two extremely valuable sources for directories of business, government,
and associations. These arc Gale Research and U.S. West Direct, both with catalogs
on the web. Tf'you can't find a directory pertaining to vour search from cither of
these sources, it probably doesn't exist. U.S. West even offers to find a relevant
directory for you if they don't already carry it.
THE DIRECTORY SOURCE
U.S. West Direct
c/o PDC
13100 East 39th Avenue, Unit U
Denver, CO 80239-3527
U.S. and Canada, phone: 1-800-422-8793, Ext. ISO
Other countries diaJ international access code, then: 1-303-375-0707, Ext. 160
U.S. and Canada, fax: 1-800-522-8793
Other countries fax by dialing your international access code, then: 1-303-3739878
h ttp: / / www. w3, us wes t. com / di re c lorysou rce /us w_di rec tory. h tin 1
Call tor pricing and availability as both are subject to change.
GALE RESEARCH
835 Penobscot Building
645 Griswold Street
Detroit, MI 48220
Phone: 313-961-2242 or ttOO-877-GALE
Fax: 313-961-6815
h tip: //www.gale. coin /
h ttp: / /www .thomson.com/gale/ddault.html

The Yellow Bonk Leadership Directories arc another excellent scries of directories
of key personnel in the U.S. federal, state, and local government; foreign business
and government representation in the^U.S.; and U.S. corporate, legal, financial,
association, and news media fields.
YELLOW BOOK LEADERSHIP DIRECTORIES SERIES
Leadership Directories, Inc.
104 Fifth Avenue, 2nd Floor
New York, NY, 10011
Phone:212-627-4140
Fax: 212-645-0931
$190-$265 each; $1,460-$1,600 complete library
A couple of good collections of web-based phone and fax listings are Kapitol and
the Reference Shelf. Kapitol includes a number of more obscure linkssuch as
the fax directory of Sierra Leoneand rates all their links by ease of use, completeness, and reliability.
255

Appendix

256

KAPITOL
31 Rue E. Gossart
B-1180 Brussels - Belgium
Phone: +32-2-344-42-42
Fax: +32-2-344-22-96
E-mail: kapitol@infobel.be
THE REFERENCE SHELF
http: / /www .synapse.net/ -radio/ refe re nc .h tni
KapitolInternational Directories
ht tp: / / www. infobel.be / i nfobel / i n fobclworld. h tml
Most military contacts can he located through two sources: the U.S. military's
Defense Link on the web, or Jane's annuals, available in prim or on CD-ROM.
DEFENSELINKhup://www.dtic.dla.mil:80/defenselink/index.htnil
Links to nearly all publicly accessible U.S. military web sites
JANE'S ANNUALS AND JOURNALS
http: / / www.j an es.com/janes.html
http: / /www. thomson.com/janes/default.html
JANE'S CANADA
Van we 11 Publishing, Ltd.
1 Northrup Crescent
P.O. Box 2131, Stations
St. Catherines, Ontario L2M 6P5
Canada
Phone: 905-937-3100
Fax: 905-937-1760
JANE'S UK
Sentinel House
163 Brighton Road
Coulsdon, Surrey
CR5 2NH
UK
Phone: 44-181-700-3700
Fax: 44-181-763-1006
JANE'S U.S.
1340 Braddock Place, Suite 300
Alexandria, VA 22314-1651
Phone: 703-683-3700
Fax: 800-836-0297
JANE'S WEST COAST
17310 Redhill Avenue, Suite 370
Irvine, CA 92714
Phone: 714-724^868
Fax: 714-724-1576
Besides the information available from the directory clearinghouses listed above,
the U.S. government is well represented on the web. A comprehensive starting

Appendix

257

point can he found at the FedWorld site, while the individual branches of government have excellent sites of their own.
FEDWORLD h tip://www.fedworld.gov/
OFFICE OF THE PRESIDENT http://www.whitehouse.gov/
U.S. HOUSE OF REPRESENTATIVES http://www.house.gov/
U.S. SENATE http://www.senate.gov/
The Foundation Center is an excellent source of directories arid information
about U.S. and international foundations.
THE FOUNDATION CENTER
79 Fifth Avenue, 8th Floor
New York, NY 10003-3076
Tel: 212-620-4230
Fax; 212-691-1828
http://fdncenter.org/
In addition to the directories available through Gale Research and U.S. West, contact information for foreign governments, businesses, finances, and associations
can be found at the sources listed below.
EUROPEAN UNION
Delegation of the European Commission to the United States
2300 M Street, NW
Washington, DC 20037
Phone: 202-862-9500
Fax: 202-429-1766
http: / / www. eurunion .org/
INTERNATIONAL MONETARY FUND (IMF)
External Relations Department
Washington, DC 20431
Phone: 202-623-7300
Fax: 202-623-6278
h ttp: / /www .irnf.org/
KR ONDISC Worldwide Government and Defense Directory on CD-ROM
$2,900 (includes 4 quarterly updates)
(Includes: Worldwide Government Directory, Worldwide Directory of Defense Authorities,
and Profiles of WorUuride Government Leaders.)

Knight-Ridder Information, Inc.

2440 El Camino Real
Mountain View, CA 94040
Phone: 415-254-8800
Fax: 415-254-8123
E-mail if you are located in the United States; customer@corp.dialog.com
E-mail if you are outside the United States: ind@corp.dialog.com
UNION OF INTERNATIONAL ASSOCIATIONS (UIA)
Rue Washington 40
B-1050 Brussels

258

Appendix

Belgium
Fax: 32-2-646.05.25.
E-mail: weblist@uia.be.
hitp: / /www. uia.org/
Tens of thousands of organizations listed on-line, with contact points. The UIA
also offers the four-volume Yearbook of International Organizations in print and
CD-ROM versions. Print version: USf 875.
THE UNITED NATIONS
New York, NY 10017
Phone: 212-963-1234
http: / /www. un.org/
THE WORLD BANK GROUP
1818 H Street, NW
Washington, DC 20433
http: / /www. wo rldban k. org/
Web site contains contact phone numbers by topic,
WORLDWIDE GOVERNMENT DIRECTORIES, INC.
7979 Old Georgetown Road
Bethesda, MD 20814
Phone; 301-718-8770 or 800-332-3535
Fax: 301-718-8494
h tip: / / www. wgd .com/
1,450 pages, $347
There are a Urge number of trade and business directories available in print and
electronic form. Many of these are available from Gale Research or U.S. West.
Other excellent sources of information include Dun & Bradstreet's large number
of publications and services and the Thomas Register. International Thomson
Publishing, which owns Jane's and Gale Research, also offers a large number of
CD-ROM, on-line, and diskette directories related to international business.
DUN & BRADSTREET INFORMATION SERVICES
3 Sylvan Way
Parsippany, NJ 07054-3896
Phone: 201-605-6000 or 800-234-3867
Fax; 201-605-6930
h ttp: / /www. dnbcorp.com/
DUN & BRADSTREET, LTD.
Ho liners Farm Way
High Wycombe
Buckinghamshire HPI2 4UL
England
Phone: 44-494-423-600
Fax: 44-494-423-595
DUN & BRADSTREET ONLINE SOLUTIONS http://www.dbisna.com/
INTERNATIONAL THOMSON PUBLISHING
Informalion/Reference Group
835 Penobscot Building

Appendix

259

645 Griswold Street

Detroit, MI 48226
Phone: 313-961-2242 or 800-877-4253
Fax: 313-961-6083
h ttp: //www. Thomson. com /
THOMAS PUBLISHING CO.
5 Penn Plaza
New York, NY1000I
Phone (US/Canada): 212-290-7277 or 800-699-9822, ext. 444
Fax (US/Canada): 212-290-7365
Phone (International): 212-290-7279
Fax (International): 212-290-8825
h ttp: / /www. ihomasregister.com/
The Thomas Register of American Manufacturers, Products, and Services {33 volume set;
annual)
Includes subsets on Products and Services, Company Profiles, and Catalog File.
Price range for set from $255.80 (in U.S.) to $495 (in Central/South America)
Thomas Register is also available on CD-ROM
Price for CD-ROM from $402 (in U.S.) to $490 (International)
INTERNET SEARCH EN6INES
We regard Internet search engines as a vitally important part of searching the net.
For that reason we are providing this list, which describes and evaluates each major
engine. We are giving you name and address, whether it searches Usenet groups as
well as the web, whether it has advanced search features that allow you 10 tailor
your search request, and are evaluating ease, content, and speed on a 1 to 10 scale,
widi 10 being the best. The evaluations are subjective, and reasonable people can
quarrel widi them, but it is our call and this is how we see it.

USENET?

1DVJUKEI
SUICH?

EASE

CONTENT

SPEED

Ahoy! (personal web page)

http ://ahoy-cs . washin gton .edu:6060

n/a

Alia Vista (general)

http://www.a ltavista.diflital.com

Keyword. Boolean
for advanced
search.

Good information source: large

number of responses makes retrieving
specific information difficult.

American Directory Assistance

n/a

Names.

Provides street maps. Entering

specific state does not allow for large
searches.

Big book (business/company finder)

http ://w ww . b igbook.com

n/a

Names.

Comprehensive listing for businesses.

Does not provide enact search option.

Bigfoot (people finder)

http ://wwW. bigfoot.com

n/a

Names, regions,
states, etc.

Major difficulty is accessing

homepage and search engine:
sometimes is not available.

Cosmix Insane Search (metal

http://www.cosm Ii.com/motherlo3d/insane

Boolean. Menudriven function.

User chooses how to search. Results

ara compiled and relevancy is better
than most "meta" search engines.

Deja News Service (newsgroup)

http ://ww w. d ejans ws. com

Keyword or
newsgroup
classification.

Nice, updated directory of newsgroup

postings.

Dopgpile (meta)
http://ww w. dogpi le.com

Keyword. Boolean
searching,

Handy time delimiter gives user some

control. User must search within

SEMCN ENiIHE HIME, TYPE, ADDRESS

INPUT

COMMENTS

Names.

First and last name muat be entered.

Missing data that can be found in
other location a.

(people finder)
http://w ww .abii.com/lookupusa/ada/ada. htm

individual engines.
Excite (general)

http ://www.excite. com

Four 11 Directory Services

(people finder)
http://www.four! 1 .com

n/a

GTE Yellow Pages (business finder)

http ://superpages. gte. net

n/a

Keyword. Boolean
for advanced
searching.

Names, countries,
states, cities,

Advanced search capabilities are

tough to learn but are extremely
powerful.

E-mail search is very good. Phone

number list is not as comprehensive,

domains.
Keyword. Boolean
and menu-driven.

Lots of graphics makes loading very

slow. Often returned no matches.

Galaxy (general)
http://galaxy.tra dawave.com/galaxy. htm I
motherload/insane

Keyword. Boolean Allows Gopher and Telnet searching

is clickable option. as option*. Provides a fairly lengthy
excerpt from page.

Highway 61 (meta)
http://www.highway61.com

Boolean is menudriven function.

Aims at being humorous; could be

more straightforward and well
defined.

HotBot (general)
http://www.hotbot.com

Boolean Is menu driven function.

Nifty searching options. User can

search for people, phrases, etc. Also
a good date delimiter.

Infohiway (general)
Http ://ww w. i nfohi way. com

Keyword.

Confusing search interface will

frustrate novice web users. Irrelevant
hits sometimes result.

Infoaeek (general)
http://ww w. infoseek.com

Natural language
or keyword.

Subject searching is available. Natural

language (unctions are a nice frill but
of little use.

Infoapace People Search (people finder)

http://in-101.infospace.com/iui/people.htm

n/a

Names,
addresses.

Comprehensive, versatile people

finder with phone, address, and
e-mail capability.

Internet Address Finder (people finder)

http'//www. iaf.net

n/a

Names,
organizations,
domain name,
or e-mail
address.

Allows searching for user by e-mail

address and by name. Available in
Dutch, French, German,
Italian, Portuguese, and English.

O)

USENET?

AOVANCEI
SEAHCN?

EASE

CONTENT

SPEC!

n/a

Looks mart (general)

http ://ww w. looksmart.com

Lycos (general)
http: //www. lycos, com

Magellan (general)
http.7/www .mckinley.com

Meta crawler (meta)

http: //www, metacrawler.com

10
SEARCH ENGINE NAME, TTFE, ADDRESS

INPUT

COMMENTS

Keyword with
menu-driven
functions for
search refining.

Hits are comprehensive and

informative. However, speed is not
impressive.

Category search
drives hits.

Java applets make loading the page

and search slow. Must search by
category.

Keyword.

Sounds and pictures are searchable.

Irrelevant hits sometimes retrieved.

Keyword.
Boolean with
menu -driven
search.

Allows subject searching; particularly

low on content for a general search
engine. Good advanced options.

Keyword and
menu-driven
options.

Combined hits are pooled together,

though relevancy is well below
average.

Open Text (general)

http://lndex.opentext.net

Keyword.
Boolean

One of the more relevant search

engines. Allows refining of search
after search is made to find more
specific items on the web.

Pathfinder (general)
http: //www. pathfinder.com

Keyword with
clickable options.

Graphics on search interface make

load time slow.

Reference (newsgroup)
http ://w ww .reference.com

Keyword with
Boolean searches.

Relevance very poor. Updated often,

but hits are very erratic.

Linkstar (people finder)

http :// www. linkstar.com

iiMfcfc___
^^BHMi^A^^^^^^^^^^^^^^^^^^^^^^^^^^^^
pjPpP^^B^Pp^gppa^pg^^^^g^^$^l|iulw..jAjm w
HP^BJ^WP^^Bi^^^W^^^^WrBWJW111-- -

SawySearch (mrta)
http://guatalili.cs.colostate.edu: 2000/f orm

Switchboard: Find a person (people finder)

http://www. switch board.com

n/a

Tilanet (newsgroup)
http: //tile.n et/ne ws/search . html

WWW Worm (general)

hltp://ww ww. cs. Colorado, adu/w www

Who Where (people finder)

http://ww w. whowhere. com

n/a

Worldpages Power Find (people finder)

http:///ww w. worldpages.com

n/a

Web Crawler (general)

www. webcra wle r.com

555-1212 (people finder)

http://www.555-12l2.com

OJ

Keyword, Boolean
is menu-driven.

Results are categorized by search

engine. No relevancy ranking, but
integration Is an option. Retrievals
available in 23 languages.

Name and
category driven.

Allows user to find business by

category and people (latter listing
is not well updated).

Keyword.
Boolean.

Primitive searching capabilities search

from a small number of newsgroup
postings.

Yahool (general)
hp;//www. yahoo, com

n/a

Keyword. Boolean
is menu -driven.

Limited capabilities; content does not

match other general engines.

Boolean, keyword
searching.

Easy-to-use general search engine.

Well laid out and user-friendly.
Relevancy is suspect.

Names.

Good e-mail finder with one weakness:

Only identification is through e-mail.
Refers user to other services to find
people.

Keyword.

Alows subject/category searching and

people search. Extensive list of city
maps.

Name, location,
domain name.

Allows searching of e-mail addresses,

telephone numbers, and web sites by
area code.

Keyword o r
location search.

Business, people, and government

searches are allowed. U.S. locations
must include state entry.

265

Appendix
MONTHLY ON-LINE USA6E COST
O

Based on 10 hours of on-line time, 100 search queries, and 100 document downloads or prints.

1-

cc

m
S

'-

*f

rt

tO

r**

5-

S
O
1

(D
ri

0?
o!>
?

S
o
r^

fn

ffi

rt
rJTJJ co co ft r3
0
6
6
0
^ o o o o

415-254-7000

DZ

214-951-7171

uj

UJ
CO

11 1

in

OJ

CO

CO

Csl

6
(O

6
o
CD

in
CM

ey

Q!

0*

PI
(O
N
0
J*1!

in

CD
5
PI

in

6 6
6
o S
o

u
K
S
o
o

o3
fl
u

ij

$19.95 flat rate

$24.95. 31.95 per hour after first 20
$9.94 flat rate
S'9.95 flat rate

Dialog*
Dow Jones*
EyeQ
FT Profile
LexU-Nexis (per use)
Lexis-Nexis (flat rate)
Profound
Pro quest Direct*

TIME
$530
free
$39

$240
$360
$4250 (flat rate!
$898

DOWI*
DOWNLOADS/SEARCHES
TOTAL
$407

$937

$1121

$1121

$450*
500
$4000
$200
free
$383

$489
740
$4360
$4450
$898

c
^ u

.c

_c

01
'S

S
C ft

UJ ? S | i S
Z

UJ

e
'-^

S I1 t t f
^

-E

-f

a.

"

* fS
t
*
j

u "3,
S"
umi.com

-^

o
=

i s t

J i s
-? -f -?

n'S

CL

fl- * t

!9
C K
LJ

=5 "S
eg qj

"
d t*-i
o
i

D
. ii
1
.s
*

o
UJ

UJ

FT Profile

Dow Jones

_^

Dialog

- S * 8 > r
3 s s" "a S
i 1 5 E 1 3

1*3
1 e1 e*
S
_l

""Includes charges on full-lexi documencs viewed hut not downloaded.

Q-

O.

LIBRARIES WORTH VISITING ON-LINE

The sample below was determined on ease of use, content, and updates of information. When dealing with foreign libraries, most especially in Europe, there is a
serious lag problem that reduces the site's usefulness. These are lihraries that we
like to use.

CJ

"E j:

k-

"Service charges vary depending on source accessed. Esdmates are based on averages taken
from a representative sampling of sources.

o 5
-a

profound.com

--^

-a
,
^.

lexisnexis.com

0
CB

2
E

Tl

info. ft. com/o nline/|

o
"

i
?

inf onauti cs. com/en

Z
f
^

?
I

(Q

01

data times .co m/proi

<

S C
S '-

i 1I

o
tJ

5 "u
*^

.f
S

) wj ones . co m/on I i ne

dialog.com

a
u
tn
u
z
j
B
z
o D

264

TIME (includes docuilienl downloads, prlnla. tf

HIGH-END ON-LINE
SERVICES

^
(A
U
U

LOW-END ON-LINE SERVICES

America Online
CompuServe
E Library
Microsoft Network

H" 3

a. a

v qj
W 3

United States
BRIGHAM YOUNG UNIVERSITY
Telnet: library.byu.edu
,
128.187.11.10
NOTE: Select vtlOO terminal and select "a" on Gateway menu.
Has unlimited number of returns, plus full citation.
GEORGETOWN UNIVERSITY
Telnet: library.Iausys.georgeiown.edu
141.161.93.5
NOTE: Log-in is george.
Access law and medical library, but limited display of information.
HARVARD UNIVERSITY
Telnet: hollis.harvard.edu
128.103.60.31
NOTE: At prompt, lype hollis.
One of die bestplus you can e-mail your results back to yourself.

266

Appendix

JOHNS HOPKINS UNIVERSITY

Telnet; janus-gate.nise.jhu.edu
128.220.8.2
NOTE: Log-in isjanus.
Access the Woodrow Wilson Collection and the American Psychology Association
database. Full display of records.
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
Telnet: Library.mit.edu
18.92.0.26
NOTE: Ix>g-in is library, Password you press enter.
Clumsy search but great library.
PRINCETON UNIVERSITY:
Telnet: Catalog.princeton.edu
128.112.204.71
PURDUE UNIVERSITY
Telnet: thorplus.lib.purdue.edu
128.210.21.22
NOTE; Log-in is cauiet.
Access to other regional libraries.
RICE UNIVERSITY
Telnet: alexandria.rice.edu
128.42.74.100
NOTE: Log-in -dS Guest.
U.S. NAVAL ACADEMY'

Telnet: LJbrary.nadn.navy.mil
131.121.1881
NOTE: Log-in is library.
UNIVERSITY OF CALIFORNIA
Telnet: melvyl.ucop.edu
192.35.222.222
NOTE: Set vtlOO or Help when prompted.
Access the California system. Some subject area databases are useful to
businesses. Sends result to you via e-mail.
UNIVERSITY OF NOTRE DAME
Telnet: irishmvs.cc.nd.edu
129.74.4.5
NOTE: At Enter command or Help prompt enter library.
UNIVERSITY OF TEXAS AT AUSTIN
Telnet: utxuts.dp.utexas.edu
128.83.216.12
NO re: Log-in is utcat.

Appendix
Australia

UNIVERSITY OF SYDNEY'
Telnet: Iib7.fisher.su.oz.au
129-78.72.7
NOTE: Log-in is library.
Lots of Pacific basin material.
Hong Kong
CHINESE UNIVERSITY OF HONG KONG
Telnet) library.cuhk.edu.hk
202.40.216.17
United Kingdom
CAMBRIDGE UNIVERSITY
Telnet: nl.cam.ac.uk
131.111.164.59
OXFORD UNIVERSITY
Telnet: library.ox.ac.uk
129.67.1.46

NOTE; At die terminal type enter vtlOO.

267

INDEX
Actionable intelligence, 50, 54, 109
Active intelligence, 77
cost of, 30, 164, 166, 167, 168
and gossip, 166-67, 175-82, 184
and human intelligence, 169,
171-83
and legal issues, 242, 243-53
limitations of, 183-86
and location of information, 169,
183
and mission, 169, 183
and offering money, 180
open ended, 184, 185-86
operational aspects of, 168-73
passive intelligence compared with,
147, 166-67, 182-83
and public domain, 184-85
right person for, 172-82, 183
and security, 166, 177-82, 185
and technical intelligence, 169-71,
172
uses of, 164, 183-86
See aki> Espionage; Human
intelligence
America Online (AOL), 14, 119, 137,
138
Analysis, 2, 7,42, 168, 169
art and craft of, 188-91
customers' role in, 187-88
and expcro, 203-4
financial, 205-6
as fortune-telling, 193, 204-12
and gathering information, 199,
203
of intenial resources, 112-13
and issue-driven projects, 200
and knowledge curve, 190, 191
methods of, 198-202
of operating environment, 200201
operational intelligence
distinguished from, 23-24
operations versus. 188
and personal relationships, 210-12
purpose of, 188, 189, 191
and searching for unanticipated,
201

and situational awareness, 189-91,

192-93, 194
and size of business, 192
as step in imelligence process,
79-81
See also Analysis
Analysts
businessmen as, 188, 192, 213
characteristics and skills of, 191-97,
212-13
financial, 205-6
as fortune-tellers, 193
issues fating, 202-5
leaks by, 233
and personal contacts, 194-97
tasks of, 23-24
types of, 197
Arab oil embargo, 195-97
Archives, 81-82, 109, 138, 139, 140,
205, 225-27
Assumption inventory, 198-200
B Team approach, 86, 103, 199, 213
Binary system, 37-38
Bookstore case, 207-10
Bribery, 245-46
British Airways, 101-4, 251-52
British intelligence, 17, 19, 20, 32-33,
39
Bulletin boards (BBS), 121-22, 127
Bush, George, 22-23
Business intelligence, 27, 42, 63
core problem of, 45
holy trinity of, 43-55
national intelligence agencies
compared with, 17
organization of, 28
purpose of, 50
tracking of. 28-31
Businessmen, as analysis, 188, 192,
213
Cataloging systems. 140-41, 142
CD-ROMS, 116-21, 142
Central Intelligence Agency (CIA)
analysis by, 115,213
and coumerintelligence, 213

Index

270

Central Intelligence Agency (CIA)

(emit.)
databases of, 40, 115
directors of, 22-23
and electronic intelligence, 32, 36
as espionage agency, 2-3, 39
former agents of, 218
honesty of, 20
and industrial espionage, i?17-19
information management at, 2-3
and knowledge crisis, 2-3, 6-7
knowledge curve of, 54
mission/function of, 6-7, 18, 19
as model for business, 17
and passive intelligence, 114-16
and pizza count, 165
and political warfare, 25
and secret information, 115-16
Chinese intelligence, 217
Classified information, 24415
Collection of information. Sef
Gathering information
Collectors of information, 96-101,
112, 113
Colorado Alliance of Research
Libraries (CARL), 139
Competitive intelligence, 7
CompuServe, 119, 137, 138
Computers, 40, 113, 228
binary system for, 37-38
and counterimelligence, 216-17
and electronic copying, 239-41
as electronic typesetting, 37, 38-42
and knowledge crisis, 5-6, 12-13,
14
and legal issues. 249-50
moderns for, 119-20
and public domain access, 37-42
as workhorses of intelligence
system, 118-19
See aiso Databases; Digital
information; Electronic
intelligence; On-line services
Consultants, 84, 85-87, 93-94
Contact records, 163
Continual monitoring curve, 68
Copyrights, 238-43, 245
Coun term te 1 ligen ce
cost of, 217,220, 221,232
and digitization, 214, 217, 219-20,
221-22, 224-32

and electronic intelligence, 224-27,

228-29
and employees, 227-28, 233
and encryption, 217
and gossip, 233
and Internet, 232
and leaks, 224-26, 227, 233
and national intelligence agencies,
215-16,217-18
and organized crime, 219
and press releases, 226-27
and proprietary information,
214-16, 244'
and secrecy, 220-24, 233
and security, 220-24, 229-32
and software piracy, 241
and technology, 233
and tracking, 233
Covert operations, 24-26, 114, 176
Cuban missile crisis, ]t>5-66, 167
Customers, 21-22, 27, 102-3, 187-

88
Data, 4, 5, 252. See aha Databases
Data Corporation, 41^42
Databases, 14, 71, 106, 113, 115
as archives of information, 81-82
and binary code, 38
of British Airways, 103-4
government, 138-39
law about, 242-43
and public domain access, 38-42
searching of, 38-42, 47, 72
Sef ahn On-line services; specific
database
Defense Technical Information
Center (DT1C), 73-74, 138-39
Dialog Information Services, 39, 40,
42, 73, 138-39
Diffusion of information, 87-94, 112
Digital information, 117, 141, 214,
217, 219-20, 221-22, 224-32
Direct information, 109
"Discovery", 237, 238
Discrete operations curve, 67-68
Disinformation, 159, 168, 226
Dissemination of information, 7,
87-94,96, 109-10, 111-13
Dissipation of information, 87-94, 95
Dow Jones news service, 39, 133-34,
135, 138

Index
E Library, 134-35
E-mail, 119, 137, 230, 231, 232
Economic Espionage Act (1996), 236,

248
Efficiency, 90, 91, 93, 104-5, 163,
223-24, 235
Electronic copying, 239-41, 249-50
Electronic intelligence (zone 1)
as category of technical
intelligence, 31-33
chart about, 48
and counterintelligence, 225-27
efficiencies of, 55
and espionage, 3435, 36
examples of sources for, 143-45
and information management, 32,
33, 35-36
as information zone, 47, 55
and knowledge curve, 54-55
and legal issues, 254
management of, 32, 33
and national intelligence agencies,
31-33
open source, 116-20
and passive intelligence, 30, 158,

254
searching for, 33, 38-42, 139-40,
141
and semiacdvc intelligence, 142,
144, 158
storage of, 33, 36-39
Set also Computers; Databases;
Internet; specific database
Electronic typesetting, 37, 38-42
Employees
and countermtelligenee, 227-28,
233
indoctrination of, 101-4
.,
nondisclosure agreements for, 249
See also Internal resources
Encryption, 217, 230, 232, 244, 250
Espionage, 1-2, 27, 49, 91, 238
and electronic intelligence, 34-35,

36
and human intelligence, 31, 33-34
as legal issues, 186, 235-38
See also Central Intelligence Agency
(CIA); Counterintelligence;
Covert operations; National
intelligence agencies
Executive News Service (ENS), 137

271

Experts, 161, 203-4, 218-19

EyeQ, 134, 136
Federal Bureau of Investigation
(FBI), 19,20
FIDOnet. 121-22
Food Lion case, 184-85
Food-processing company case,
151-57
Fortune-telling, 43-45, 55, 193,
204-12
Frank, Jacqueline, 165
Freedom of Information Act, 245
French intelligence, 217, 218
FT Profile, 39, 136,225
Gathering information/intelligence
and analysis, 199, 203
collectors for, 96-101
by journalists, 238
and knowledge generation cycle, 7
by lawyers, 237, 248
and legal context of intelligence,
235-54
operational, 23-24
as step in intelligence process,
72-75
and lapping internal resources,
85-109, 110
See also Active intelligence; Covert
operations; Electronic
intelligence; Espionage; Human
intelligence; Internet; Passive
intelligence; Semiactive
intelligence; Technical
intelligence
General Motors/Lopez case, 222-24,

228, 236
Geography. 10-11, 12, 42, 45-47, 52,
69-70, 144
Globalism, 3-8, 253-54
Gossip (zone 3), 9, 72, 233
accessing, 52
and active intelligence, 166-67,
175-82, 184
and diffusion of information, 88-90
as information zone, 48, 52, 54
and internal resources, 159
and knowledge curve, 54
and semiactive intelligence, 142,
158-59. 160

272

Gossip (zone 3) (conl.)

and tapping internal resources,
88-90, 96
Gray area (zone 4), 48-49, 52, 54, 71,
116, 158, 159
Gulf War, 165
Human intelligence, 27, 31, 33-35,
95, 169, 171-83, 207-12. See aho
Internal resources
lACInsite, 135-36
IBM problems, 214-15
Indirect information, 109
Industry links (Lycos), 128
Info lust, 148-49 '
Informal intelligence systems, 8-15,
148. See also type of system
Information
availability of, 3-8, 27, 29
definition of, 4
expansion/overload of, 3-8, 84-85,
101, 111, 116, 189
life cycle of, 50
requesting, 244 46
significance of, 50, 87-S8, 89, 90,
92, 95-96, 105-6, 109, 111, 166
value of, 51, 92-93
See alia type of information or
intelligence
Inlbrmationinventory/re-inventory,
65, 66-67, 75-76, 151
Information management, 2-3, 6, 32,
33,35-36,42,95, 111-13
Information Requirements sheet,
66-67, 151
Information zones, 47-54, 70-72
See also specific zone
Infoscck, 124-25, 142
Integrated Services Digital Number
(ISDN), 119-20
Intellectual property. See Copyrights:
Trade secrets
Intelligence
connotations of, 1-2
foundations of, 56
honesty of, 20
and knowledge crisis, 6
privatization of, 3
as probability, 153
security as separate from, 19-20

Index
as specialization, 78
See aho Espionage; type of intelligence
Intelligence agency, definition of, 6
Intelligence directors, 6, 7, 22-23, 27,
62
Intelligence process
chart of, 57, 58
computers as workhorses of,
118-19
cost of, 100-101
first task in building, 1011
as generic, 82
and geography of information,
69-70
goal of, 61
honesty as core of, 14-15
informal, 8-15, 148
information inventory/re-inventory
in, 65, 66-67, 75-76
integrity of, 57
and knowledge curve, 67-70
magnetic water conditioner case as
example of, 58-82
mission of, 107
overview of, 82-83
project as distinct from, 82-83
and size of business, 13
steps in, 60-82
Intelligence projects, 53, 82-83
Intelligence specialists, 56-57
Internal resources, 148, 159
chart about, 93, 110
and collectors of information,
96-101, 112, 113
cost of tapping, 94
and diffusion/dissipation of
information, 87-94, 112
and dissemination of information,
88, 109-10, 111-13
evaluation of, 112-13
and gathering information, 85-109
and knowledge curve, 109
location of information within, 98,
104-9
management of intelligence from,
95, 111-13
and storage of information, 109,
110, 113
and users of intelligence, 111-12
Internal Revenue Service, U.S., 32
Internet, 14, 144,232, 241

Index
applications chart for, 123
content of, 120, 129-30
cost of, 129-30
and electronic intelligence, 47,
116-21
and legal issues, 249-50
libraries accessible on, 139, 141
modem as connection to, 119-20
and on-line services, 137, 149
origins of, 120-22
and passive intelligence, 120-32
and semiactive intelligence, 158,
159
using the, 123-30
See also World wide web
Intimacy. See Informal intelligence
systems
Intuition. See Informal intelligence
systems
Iran, personal contacts in, 210-11
Israeli intelligence, 17, 19, 20, 218
Issue-driven projects, 200
Joseph (biblical person), 43-44, 187,
193
Kinko case, 239
Knowledge
computers as means to, 5
crisis of, 1-15, 84
definition of, 45
domains of, 47-50
as power, 20
See also Knowledge curve; specific
information zone
Knowledge curve, 50-55, 67-70, 109,
148, 190, 191
Knowledge generation cycle, 7
Knowledge managers, 6, 7, 22-23,'27,
62
Laker Airways, 251-53
Leaks, 166, 177-78, 185, 221-22,
224-26, 227, 233
Legal issues
and active intelligence, 242, 243-53
and bribery, 245-46
and classified information, 244-45
and computers, 249-50
in context of intelligence, 235-38
and copyrights, 239-43, 245

273

and data/databases, 242-43, 252

and electronic copying, 23941,
249-50
and electronic intelligence, 254
and encryption, 244, 250
foreign, 235, 236-37, 244, 246, 247,
250-53
influences for changing, 253-54
and Internet, 249-50
and passive intelligence, 238-43,
253, 254
and photocopies, 238-39
and proprietary information, 245,
249
and requesting information,
244-46
and semiactive intelligence, 242,
243-53
about software, 241-42
and taking information, 24750
and tape recording, 24617
and transnational information,
250-53
and wiretapping, 246-47
Levchenko, Stanislav, 214-15
I-exis, 39, 41, 42, 59, 131-32, 133,
134, 135-37, 139, 155, 225, 227,
240
Libraries, 51, 70-71, 74, 116,131-32,
135, 139-41, 142
Library of Congress, 141
Lopez de Arriortua, Jose Ignacio,
222-24, 228, 236
Magic box, 91, 94-95, 101, 104,
109-10
Magnetic water conditioner case
acquiring information in, 7275
analysis in, 79-81
archives and database for, 81-82
background information about,
58-60
knowledge base in, 64-67
knowledge curve in, 67-70
mission in, fiO-63
overview of process for, 82-83
recommendation in, 80-81
retasking system for, 75-79
targeting zone in, 70-72
Mail-room clerk case, 87-88, 89, 90,
92, 95, 105

274

Market Insite, 135-36

Mass-market on-line service, 137
Mead Corporalion, 41
Mead Data Central, 41, 42
Metasearch, 125-28
Microsoft, 123, 139, 216, 217
Mission
and active intelligence, 169, 183
and building intelligence system,
101-4
of CIA, 6-7, 18, 19
of intelligence system, 107
of knowledge managers, 6, 7
in magnetic water conditioner case,
60-63
and semiactive intelligence, 149,
151, 152
as step in intelligence process,
60-63
Mission creep, 83
Modems, 119-20
Money, 43-55, 180- See also
Knowledge curve
Motherlode Insane Search, 129-30
MSNetwork, 139
Myhrvold, Nathan, 216, 217
National Imagery and Mapping
Agency (N1MA), 32
National intelligence agencies
business intelligence compared
with, 17
and counterintelligence, 195-97,
215-16,217-18
and covert operations, 24-26
customers as center of, 21-22
and databases, 115
failures of, 195-97
former agents of, 218-19
and open source intelligence, 26
organization of, 18-26
purpose of, 21-22, 54
and technical intelligence, 31-33,
169-71
See aim specific agency or foreign
nation
National Security Agency, 31-32, 33,
37-38, 40, 169-70
National Technical Information
Service (NTIS), 73, 138-39
Netscape, 122, 123-24

Index
Networks, 34, 160-63
New company contact case, 105-9
New Hampshire company case,
175-82
New York Times, 37, 38
News services, 136-37
Newsgroups, 127-28, 129, 158
NEXIS, 39, 41-42, 59, 131-32, 133,
134, 135-37, 139, 225, 227, 240
Nondisclosure agreements, 249
On-line services, 116-21, 130-39, 142,
224-27, 228-29, 240-41
Online Computer Library Center,
Inc. (OCLC), 74
Open source intelligence. 26, 27, 114,
115-20. See also type of intelligence
Open Source Solutions (OSS), 114,
116
Operating environment, analysis of,
200-201
Operational intelligence, 23-26
Operations, analysis versus, 188
Organized crime, 219
Pager case, 124-29, 132-33, 134, 138,
139, 198
Paper intelligence (zone 2), 47, 54,
74, 139-40, 142, 144, 158, 254
Passive intelligence
active intelligence compared with,
29-31, 147, 166-67, 182-83
and CIA, 114-16
cost of, 30
and electronic intelligence, 116-20,
158, 254
and Internet, 120-32
legal risks in, 238-43, 253, 254
and on-line services, 116-21, 13039
and paper intelligence, 158, 254
searching for, 147
and steps in intelligence process,
76-77
and Wildman Joe case, 157
See also Technical intelligence
Passwords, 230, 232
People
and info lust, 149
and personal relationships, 194-97,
210-12

Index

and semiactive intelligence,

146-47, 149, 151-30
See oho Employees; Human
intelligence
Personal relationships, 194-97,
210-12
Phelps, Marshall C., Jr., 214-15
Photocopies, 238-39
Photographic intelligence, 31, 35-36,
170-71
Pizzas, 165-66, 167
Political warfare, 24, 27
Press releases, 226-27
"Prime Time Live" (TV program)
case, 184-85
Prodigy, 119
Profound, 136
Project, intelligence process as
distinct from, 82
Project Gutenberg, 141
Promotional material, 145-46
Proprietary information (zone 5),
49, 52, 54-55, 142, 214-16, 244,

245, 249
Proquest, 136
Public domain, 36-42, 142-51,
184-85, 245
Quark Research Group, 247
Querying, 110, 113, See also Searching
Recommendation, as step in
intelligence process, 80-81
Reed Elsiveer, 42
Requesting information, 244-46
Research, and espionage, 238
Retrieval of information, and tapping
internal resources, 87-90
Russian company case, 172-73
"
Satellites, 150-51, 170-71. See
also Signal Intelligence
Sawysearch, 125
Scrubbing disks, 230-31
Search engines, 123-30, 132-33, 141
Searching
and active intelligence. 147
Boolean, 125-28, 135
of databases, 38-42, 47
for electronic intelligence, 33,
38-42, 117, 118, 139-40

275

of Internet, 123-30
key words for, 72-73
in Library of Congress, 141
of on-line services, 130-39
of paper sources, 13940
and passive intelligence, 147
software for, 118
for unanticipated, 201
See also Libraries
Secrecy, 49, 114, 115-16,164-65,
220-24, 232, 233. See also
Proprietary information
Security, 19-20, 166, 177-82, 185,
224-32. See aha
Counterintelligence
Semiactive intelligence
characteristics of, 30-31
contact means in, 160
and contact records, 163
cost of, 148, 156, 161
and efficiency, 161
and electronic intelligence, 142,
144, 158
examples of sources for, 145-47,
162
and experts, 161
focusing in, 159
function of, 149-50
and geography of information, 144
and gossip, 142, 158-59, 160
and gray area, 158, 159
and info lust, 148-49
and informal networks, 148
and internal resources, 148
and Internet, 158, 159
and knowledge curve, 148
and legal issues, 242, 243-53
and location of information, 159
as mission-oriented, 149, 151, 152
and networks, 160-63
and paper intelligence, 142, 144,

158
and people, 147, 149, 151-60
and proprietary intelligence, 142
and public domain, 142-44, 145-51
step-by-step approach for, 159-60
and steps in intelligence process,
76-77
as task-driven, 149
and time, 148, 156
and tracking, 144

Index

276

Service providers, 119-20, 130-39,

240-41
SIG1NT. See Signal Intelligence
Signal Intelligence, 36-37
Situational awareness, 3, 129, 189-91,
192-93, 194
Size of business, 8-9, 111-12, 119-20,
192
Slovakian case, 158
Society of Competitive Intelligence
Professionals, 7
Software, 5, 14, 118, 227, 232, 241-12
Software Piracy Association (SPA),
241-42
South African banker case, 224-26,
227
Soviet Union
defense spending case in, 213
intelligence in, 22, 26, 214-15, 218
security in, 221
Space, 43-55
Space program, 39-40
Specialization, intelligence as, 7-8
Stock market, 205-6, 207, 243
Storage, 33, 37-39, 47, 109, 110, 113
See also Archives
Strategic reviews, 199-200
Taking information, 247-50
Tape recording, 246-47
Technical intelligence. 31-33, 35-36,
169-71,172, 177

See also Electronic intelligence;

Photographic intelligence
Technology, 10, 11-13. 233. See also
type of technology
Time, 43-55, 116, 148, 156, 201-2,
232. See a/so Knowledge curve
Tracking, 28-31, 104-5, 106-8, 144,
233
Trade secrets, 236. 241, 243-44, 245,
248-49
Transnational information, 250-53
Trends, identifying, 45
Uniform Trade Secrets Act, 248-19
Unpublished materials, 146, 157,
245
Usenet Calculator, 127-28
V&M Distributors, 126, 127
VHF Communications, 128-29
Vietnam War, 25, 34
Virgin Atlantic Airlines, 251-52
Volkswagen case, 222-24, 228
Westlaw, 41, 155
Wildman Joe case, 151-57
Windows-Intel (Wintel), 118
Wiretapping, 32-33, 246-47
World wide web, 122, 123, 129-30,
241. See abo Internet
Zero-based reasoning, 200-201