This case study taken from the text book :Management Information Systems, Managing the digital firm,

Laudon & Laudon. Copyright © Pearson

Chapter 6 Foundations of Business Intelligence: Databases and Information Management 227

I N T E R A C T I V E S E S S I O N : T E C H N O LO GY
WHAT CAN BUSINESSES LEARN FROM TEXT MINING?
Text mining is the discovery of patterns and relation- tions pinpoint key events and relationships, such as
ships from large sets of unstructured data—the kind intent to buy, intent to leave, or customer “wish”
of data we generate in e-mails, phone conversations, events. It can reveal specific product and service
blog postings, online customer surveys, and tweets. issues, reactions to marketing and public relations
The mobile digital platform has amplified the efforts, and even buying signals.
explosion in digital information, with hundreds of Attensity’s software integrated with JetBlue’s other
millions of people calling, texting, searching, customer analysis tools, such as Satmetrix’s Net
“apping” (using applications), buying goods, and Promoter metrics, which classifies customers into
writing billions of e-mails on the go. groups that are generating positive, negative, or no
Consumers today are more than just consumers: feedback about the company. Using Attensity’s text
they have more ways to collaborate, share informa- analytics in tandem with these tools, JetBlue devel-
tion, and influence the opinions of their friends and oped a customer bill of rights that addressed the
peers, and the data they create in doing so have major issues customers had with the company.
significant value to businesses. Unlike structured Hotel chains like Gaylord Hotels and Choice
data, which are generated from events such as Hotels are using text mining software to glean
completing a purchase transaction, unstructured data insights from thousands of customer satisfaction
have no distinct form. Nevertheless, managers surveys provided by their guests. Gaylord Hotels is
believe such data may offer unique insights into cus- using Clarabridge’s text analytics solution delivered
tomer behavior and attitudes that were much more via the Internet as a hosted software service to
difficult to determine years ago. gather and analyze customer feedback from surveys,
For example, in 2007, JetBlue experienced e-mail, chat messaging, staffed call centers, and
unprecedented levels of customer discontent in the online forums associated with guests’ and meeting
wake of a February ice storm that resulted in planners’ experiences at the company’s convention
widespread flight cancellations and planes stranded resorts. The Clarabridge software sorts through the
on Kennedy Airport runways. The airline received hotel chain’s customer surveys and gathers positive
15,000 e-mails per day from customers during the and negative comments, organizing them into a
storm and immediately afterwards, up from its usual variety of categories to reveal less obvious insights.
daily volume of 400. The volume was so much larger For example, guests complained about many things
than usual that JetBlue had no simple way to read more frequently than noisy rooms, but complaints of
everything its customers were saying. noisy rooms were most frequently correlated with
Fortunately, the company had recently contracted surveys indicating an unwillingness to return to the
with Attensity, a leading vendor of text analytics hotel for another stay.
software, and was able to use the software to analyze Analyzing customer surveys used to take weeks,
all of the e-mail it had received within two days. but now takes only days, thanks to the Clarabridge
According to JetBlue research analyst Bryan software. Location managers and corporate execu-
Jeppsen, Attensity Analyze for Voice of the Customer tives have also used findings from text mining to
(VoC) enabled JetBlue to rapidly extract customer influence decisions on building improvements.
sentiments, preferences, and requests it couldn’t find Wendy’s International adopted Clarabridge software
any other way. This tool uses a proprietary technol- to analyze nearly 500,000 messages it collects each
ogy to automatically identify facts, opinions, year from its Web-based feedback forum, call center
requests, trends, and trouble spots from the unstruc- notes, e-mail messages, receipt-based surveys, and
tured text of survey responses, service notes, e-mail social media. The chain’s customer satisfaction team
messages, Web forums, blog entries, news articles, had previously used spreadsheets and keyword
and other customer communications. The technol- searches to review customer comments, a very slow
ogy is able to accurately and automatically identify manual approach. Wendy’s management was looking
the many different “voices” customers use to express for a better tool to speed analysis, detect emerging
their feedback (such as a negative voice, positive issues, and pinpoint troubled areas of the business at
voice, or conditional voice), which helps organiza- the store, regional, or corporate level.
228 Part Two Information Technology Infrastructure

The Clarabridge technology enables Wendy’s to determine how consumers feel about their brand and
track customer experiences down to the store level take steps to respond to negative sentiment.
within minutes. This timely information helps store, Structured data analysis won’t be rendered obso-
regional, and corporate managers spot and address lete by text analytics, but companies that are able to
problems related to meal quality, cleanliness, and use both methods to develop a clearer picture of
speed of service. their customers’ attitudes will have an easier time
Text analytics software caught on first with establishing and building their brand and gleaning
government agencies and larger companies with insights that will enhance profitability.
information systems departments that had the
means to properly use the complicated software, but Sources: Doug Henschen, “Wendy’s Taps Text Analytics to Mine
Customer Feedback,” Information Week, March 23, 2010; David
Clarabridge is now offering a version of its product Stodder,” How Text Analytics Drive Customer Insight” Information
geared towards small businesses. The technology has Week, February 1, 2010; Nancy David Kho, “Customer Experience
already caught on with law enforcement, search tool and Sentiment Analysis,” KMWorld, February 1, 2010; Siobhan
Gorman, “Details of Einstein Cyber-Shield Disclosed by White
interfaces, and “listening platforms” like Nielsen
House,” The Wall Street Journal, March 2, 2010; www.attensity.com,
Online. Listening platforms are text mining tools that accessed June 16, 2010; and www.clarabridge.com, accessed
focus on brand management, allowing companies to June 17, 2010.

CASE STUDY QUESTIONS MIS IN ACTION

1. What challenges does the increase in unstruc- Visit a Web site such as QVC.com or TripAdvisor.com
tured data present for businesses? detailing products or services that have customer
2. How does text-mining improve decision-making? reviews. Pick a product, hotel, or other service with
3. What kinds of companies are most likely to benefit at least 15 customer reviews and read those reviews,
from text mining software? Explain your answer. both positive and negative. How could Web content
4. In what ways could text mining potentially lead to mining help the offering company improve or better
the erosion of personal information privacy? market this product or service? What pieces of
Explain. information should highlighted?

Web mining looks for patterns in data through content mining, structure
mining, and usage mining. Web content mining is the process of extracting
knowledge from the content of Web pages, which may include text, image, audio,
and video data. Web structure mining extracts useful information from the links
embedded in Web documents. For example, links pointing to a document indicate
the popularity of the document, while links coming out of a document indicate
the richness or perhaps the variety of topics covered in the document. Web usage
mining examines user interaction data recorded by a Web server whenever
requests for a Web site’s resources are received. The usage data records the user’s
behavior when the user browses or makes transactions on the Web site and col-
lects the data in a server log. Analyzing such data can help companies determine
the value of particular customers, cross marketing strategies across products, and
the effectiveness of promotional campaigns.

DATABASES AND THE WEB

Have you ever tried to use the Web to place an order or view a product catalog?
If so, you were probably using a Web site linked to an internal corporate
database. Many companies now use the Web to make some of the information
in their internal databases available to customers and business partners.