Topic 4: Metrics

Introduction
Metrics are indices or measurement standards used to gauge the performance of a
website. Every web analysis is conducted by using a lot of different metrics, each of which
measures and assess a different aspect of a website. On its own, the result of a metric may
be indecisive. However, when put together, the results of multiple metrics can give a
complete picture of web performance

Learning Objectives
By the end of this topic you should be able to:

• Describe web analytics Metric

• Discuss the importance of metrics
• Discuss examples and importance of various Metrics

Metrics
After clear understanding of your digital business goals, then you evaluate how your
digital marketing efforts are performing. Assessing performance of your digital
initiatives requires you to either examine reports or a dashboard (a high-level summary
of data from various reports) in your web analytics tool. Web analytics reports provide
various kind of information-including traffic sources (where visitors are coming from),
site content (what content they’re interested in), audience characteristics (what types of
visitors you have), and conversions (what your visitors are able to accomplish). Most of
these reports have both a graphical element (a chart) and a data table. Within each report
you will find two key data elements that are inseparably connected and form the
foundation for all measurement: metrics and dimensions.

In web analytics, a metric is a quantitative measurement of online activity. Metrics are

always expressed as numerical values such as 3,451-page views or a 4.5% conversion
rate. A dimension is a set of attributes that categorize the numeric results into meaningful
groupings. In contrast to metrics, dimensions are always represented as textual values or
words (including dates). For example, if the data dimension were countries, the
dimension values would be Canada, Germany, Japan, and so on. Metrics and dimensions
go together because metrics measure something (numbers), while dimensions describe what
is being measured (text).
Types of metrics
There are three main types of web analytics metrics: counts, ratios, and other
calculations.
• A count is a number of items -usually whole number e.g 10, 800 orders.
• Ratios are calculations of two amounts in order to show the relationship between
them—most commonly dividing one count by another. Most of the ratios used in
web analytics are actually rates, which are a special form of ratio where the two
amounts have different base units (page views per visit).

Advantages of ratios
They compensate for fluctuations in volume, making it easier to compare the relative
performance of different items. For example, campaign A generated 300 orders from
10,000 visits (3% conversion rate). A smaller Campaign B created only 50 orders from
1,000 visits (5% conversion rate). Even though Campaign B had a lower volume of visits
it converted them at a higher rate than Campaign A (5% instead of 3%). You can’t ignore
the raw counts for orders (300 versus 50), but the rate is often useful in side-by-side
comparisons.

• The third type of metric is any kind of mathematical or statistical calculation that
is not a simple ratio, which may involve multiple counts and other operations
besides just division. An example of a metric that is a calculation is (Visits-
Entries)/Visits. Majority of metrics you will use in your web analytics tool will be
just counts and ratios. Calculations are part of the complex web analytics analysis
 while counts and ratios are sufficient for most business users to start deriving
insights from web analytics.

Common Metrics

Web analytics uses common metrics and most of the metrics appear by default in many
reports, therefore you need to understand what they mean so that you can interpret them
correctly.
Page Views
A page view (or pageview) represents an instance of a page being loaded and viewed in
a web browser. From a business perspective, page views indicate how much overall web
content is being consumed by site visitors. Users can produce multiple page views for a
single page if they navigate repeatedly to the same page during a visit. For many media
sites, page views take on a special meaning because each page view represents an
opportunity to display advertisements. It’s an old metric but widely used to understand
content consumptions.
Challenges of page views
• With introduction of interactive content (web 2.0), page views metric has become
irrelevant e.g Flash
• Page view does not apply to tracking other onsite interactions, such as
downloading files or clicking on external links.

Visits
A visit encompasses all the interactions that a user has with a website during a single
sitting or session. From a business perspective, visits reveal how popular your website is;
the more visits, the more popular it is. During a single visit, an individual might browse
several pages, view multiple videos, and perform various searches. In contrast, a visitor
could simply abandon the site after seeing one page—both count as a visit.

As an industry standard, most web analytics tools will terminate a visit after 30 minutes
of inactivity. If a visitor stepped away from her computer to answer the phone and didn’t
return for an hour, her interaction with the website would count as two separate visits
(one for before the phone call and another for when she returned).
Unique Visitors
Unique visitors (or visitors) are the inferred users who visited a website during a specific
reporting period. From a business perspective, unique visitors show the audience reach
of your website. When an individual visits a website more than once during a reporting
period (daily, weekly, or monthly), she will be counted as only one unique visitor. When
a visitor first comes to a website, the web analytics tool uses a persistent cookie to assign
her a unique, anonymous ID that will identify her if she returns to the site.

With unique visitors, we infer each unique visitor is a unique individual; however, in
web analytics a single individual can be seen as multiple unique visitors or multiple
people can be mistakenly viewed as a single unique visitor. For example, if you browse a
website from a work computer, home computer, and tablet device, you’ll be seen as three
separate unique visitors. If you visit the same site using two different web browsers, each
visit through the Firefox and Chrome browsers will be viewed as separate unique
visitors. If you delete your tracking cookies between visits, you’ll be seen as multiple
unique visitors.

Challenges of unique visitor

• Different members of a single household may appear as one unique visitor because
they visit a website from a shared computer and web browser. Despite the
challenges, the unique visitor metric gives you a reasonable estimation of how
many distinct individuals are coming to your website. The only way to get a
clearer measure of how many actual people visit your site would be to identify all
individuals through authentication (visitors identify themselves when they log
into your website), which wouldn’t be feasible in many cases. Web analytics tools
can also report how many new and returning visitors you have, which can be
helpful in understanding how acquisition or retention efforts are performing.
Average Visit Duration (Average Time Spent)
Average visit duration (average time spent) is the average length of a visit or session on
a particular site. From a business perspective, it indicates how much time visitors are
typically spending on your site, which sounds useful but is problematic for a couple of
key reasons.
• Average time spent includes only the time spent on the site for all of the pages
except the visitors’ exit pages. Web analytics vendors use the timestamp (a
combination of date and time) of each new page request to calculate the average
visit duration and time spent on each page. When you enter a landing page, for
example, a timestamp indicates when you entered the page and then when you
proceed to the next page; the web analytics tool subtracts the first timestamp from
the second timestamp to determine how much time was spent on the previous
page (the landing page). This approach is used throughout the visit to calculate
the total visit duration but runs into a problem on the final exit page: There is no
closing timestamp to understand how much time was spent on the last page.
Single-page visits and exit pages, therefore, are excluded from the average visit
duration calculations.
• How much time somebody spends on a website is difficult to interpret without
more context. For example, if you found people were spending a long time on your
site, you could interpret this as a sign that your content is valuable and highly
engaging. However, long average visits could indicate that visitors are unable to
quickly find the right content and becoming increasingly frustrated as they scour
the site for what they need. Without additional context from an onsite survey or
time expectations for a particular site (we want people to stay a long time or get
them in-and-out quickly), interpreting average visit duration can be difficult. The
average time spent per page can be a useful metric to pinpoint particular page-
specific issues, but context is still important. Unless you specifically ask visitors
through an onsite survey, you may never know why people are spending an
inordinate amount of time on a particular page.

Page views per Visit

Page views per visit represents the average number of pages viewed during a visit or
session. From a business perspective, it can show how engaged visitors are with your site
content or at least how much page content each visitor is consuming during a single visit.
Your business goals will shape how you use and interpret this particular metric.
Companies focused on lead generation will want to streamline the online application
process, so reducing the average page views per visit might be important if it increases
the number of leads. Media sites that are dependent on ad revenues will want to
maximize the page views per visit without jeopardizing the return frequency of their
visitors. Considering page views per visit in combination with the average visit
duration, you can get a sense for the typical activity levels of your visitors during a single
session and compare that to what’s expected by your organization. For example, if your
company provides extensive editorial content online but most of the visits are short in
nature, it could point to a problem that needs to be addressed (wrong content, poor
linking, bad naming convention, weak navigation).

Bounce Rate
Bounce rate is the percentage of entering visits (entries or entrances) that are single-page
visits or entering visits that leave (bounce from) the site after viewing only one page.
From a business perspective, the bounce rate reveals the effectiveness of your entry
pages. In other words, what kind of first impression are these pages having? Are your
landing pages encouraging visitors to go deeper into your site? If a particular page has a
high bounce rate of 80%, that means four out of five visitors entering the page are
bouncing and not viewing any additional pages. Typically, a high bounce rate means
something is wrong with the landing page, which could be due to a number of factors:
• Messaging and content that doesn’t match visitors’ expectations
• Poor page design or layout
• Slow loading times
• Web browser compatibility issue (perhaps it doesn’t display properly in the
browser)
• Insufficient links to other related content
• Confusing site navigation or no search options
• Weak or hidden call-to-actions (for example, call-to-actions are placed “below the
fold” of the web page where the visitor needs to scroll down to see them).

A high bounce rate can also be no fault of the actual landing page, but in fact due to a
problem created by the upstream traffic source, for example a wrong campaign message
would lead to a high bounce rate. For example, you would anticipate a high bounce rate
on a store location page because visitors search for address information and leave when
they find it. The same applies to blog posts where visitors are looking for only
information that answers a specific question and nothing more. If a landing page was
designed to direct visitors to external partner websites, you’d expect the page to have a
high bounce rate.
Bounce Rate Formula
Since the bounce rate is the percentage of visitors who only view one page on your site,
it’s calculated by dividing the total number of single one-page visits by the total number
of visitors. For example, if 100 people visit your site and 10 of them only visit one page,
then your bounce rate would be 10%.

What is the recommended/good bounce rate?

The average bounce rate is somewhere between 26% and 70%, with the optimal range
being between 26% and 40% (40% and below is generally ideal).

The average bounce rate can also change depending on the viewer’s device. Mobile
devices, for instance, have the highest bounce rate across all industries at 51%. Whereas
the average bounce rate on a desktop is 43% and the average for tablets is 45%.

If it’s over 90%, that’s a major cause for alarm but it's usually easy to decrease because
there’s something specific scaring everyone off.

How to Reduce High Bounce Rates

Bounce rates only tell you that someone landed on a web page and left it without visiting
any other page on your website. It does not tell you how some interacted with your
website
According to Jeffrey Vocell it is important to take practical steps to examine different
pieces of metrics to understand circumstances behind bounce rates. The following are
some of the steps
• Ensure your website is mobile-friendly. Mobile users account for about half of web
traffic globally.
• Look at your bounce rate based on different sources. For example, sources
directing traffic to a given page might have something to do with its bounce rate.
HubSpot Web Analytics Dashboard allows you to break down the bounce rate
according to source. Make URLs easy to read, remember, and type in. Check if the
user is not being greeted with an error message.
• Avoid other disruptions that might hurt the user experience. Things like full-
screen pop-ups, for example are annoying. You want visitors to be drawn into
your page and stay for as long as needed to convert. Avoid inbound messages that
disrupt the user experience to the point that causes the visitor to leave.
• Determine which keywords this page ranks for — and if your content sufficiently
covers those topics
Avoid misleading visitors about the site content and keywords. Match keyword to
content to ensure organic visitors get the content they expect.

Exit rate
Exit rate is the percentage of visits that terminate or exit on a particular page. From a
business perspective, the exit rate highlights potential site issues that cause visitors to
leave your site prematurely. Every visit or session ends at some point, but when or where
visitors exit can be important. In the case of a multi-step process, such as making an online
purchase, you don’t want the majority of your visitors exiting at the second step (billing
information page) in a six-step process. The exit rate is different from the bounce rate
because it focuses on all visits, not just entry traffic to web pages. For example, the exit
rate will include single-page visits (bounces) to a particular page as well as visitors who
navigate to the page and then abandon the site.

Context is equally important when evaluating exit rates for key pages. If a web page is a
logical exit point, then you would expect it to have a high exit rate. For example, an order
confirmation or thank you page that is shown after a successful online purchase would
have a high exit rate. When a page is an initial or transitional step in a multi-step process
or multi-page content series, you expect to see some attrition from step-to-step, but an
unexpectedly high exit rate may indicate a problem. For instance, a recent web design
update mistakenly removes a form field option that impedes a visitor’s ability to proceed
forward.

You might have other key pages on your site that you would prefer not to have a high
exit rate, such as your search results page or homepage. In some cases, it will be easy to
spot the problem that is causing the high exit rate (broken links, missing content), and in
other cases you may need to use onsite surveys to determine what is causing visitors to
exit prematurely. Although web analytics can answer many behavioral questions (who,
what, when, where, how), it can never answer attitudinal questions, such as why
someone is choosing to abandon a website.

Difference between bounce rate and Exit rate

Let's say you were comparing bounce rates and exit rates for a thank-you page. A high
bounce rate on that page would be kind of alarming, because that means people are only
viewing that page alone, then clicking away. Even worse, they didn't fill out a form to get
to it, which means you're losing out on conversions.

Conversion Rate
Conversion rate is the percentage of visitors (or visits) that reached a particular outcome
or performed a target action. Conversion rate refers to the number of people who
complete a desired conversion (primary or secondary) out of the “total audience” that
was exposed to that opportunity/possible action.

From a business perspective, the conversion rate displays how effectively your website
is getting visitors to do what you want them to do. Each organization’s online business
goals will define what it wants visitors to do and what represents a conversion (a desired
action, behavior, or outcome). For retailers, a conversion is a purchase or order.

Calculate the conversion rate with the formulae below

Conversion Rate= (Total conversions/Total Visitors) X 100
OR
= Number of Visitors Who Converted in the Period / Total Website Visitors in the
Period x 100
These are examples of Common Conversions for some companies:
• Application submission
• Subscriptions
• Email newsletter sign-ups
• Registering for a webinar
• Downloading resources like ebooks or white papers
• Completing a purchase
• Scanning a QR code
• Completing a survey
• Writing a review,
• Scheduled appointment
• App downloaded etc

For other companies, it could be an application submission, a subscription, a newsletter

sign-up, a scheduled appointment, or an app download.
Conversion rate is a unique metric for the following reasons:
• It is the only metric that is directly tied to measuring business outcomes.
• The metric isn’t uniform because the numerator of the conversion rate can be any
form of success (order, subscription, application, lead), which will vary across
organizations by their business goals.
 • You could have multiple conversion rates that measure different aspects of your
business. Understanding the business logic behind event tracking and conversions
will be critical to driving real value from web analytics.
How Web Metrics Work Together
All of these metrics share different insights into how your website and digital campaigns
are performing. Web analytics tools contain some of these tools to give data for your
digital measurements. Web analytics measure online activity and all of these metrics
cover different aspects of the online experience. When you combine them (e.g Page view
counts, average visit duration, bounce rate etc) the different data points provide
complementary insights that can help you understand what’s happening with your
website and campaigns. In the example below, Web analytics measures focus on three
main areas:
• Traffic volumes. Page views, visits, and unique visitors
• Visitor behaviors. Page views per visit and average visit duration
• Page-level interactions. Bounce rate and exit rate