Download Cambridge marKeTiNg COLLege gUide TO Web aNaLYTiCS NigeL

alumni
briefing
paper
5 : 2011
Cambridge
MARKETING
COLLEGE
GUIDE TO
WEB ANALYTICS
NIGEL BRADLEY
CAMBRIDGE MARKETING COLLEGES
ST JOHNS INNOVATION CENTRE • COWLEY ROAD • CAMBRIDGE • CB4 0WS
[email protected] www.marketingcollege.com
For further information:
Tel: 01954 234940
Email: [email protected]
Web: www.marketingcollege.com
Published in the United Kingdom by
Cambridge Marketing Colleges
Cygnus Business Park
Swavesey, CAMBRIDGE CB24 4AA
© Nigel Bradley 2011
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
CONTENTS
page
PREFACE
2
INTRODUCTION
2
SECTION 1 What Are Web Analytics?
4
1a. The Web Analytics Industry
4
1b. Logfile analytics
6
1c. Dashboards
7
1d. Important services today
9
SECTION 2 What Are The Opportunities For Marketers?
10
2a. Setting up your web pages for analytics
10
2b. Bounce Metrics
12
2c. Segmentation filters
12
2d. Conversion metrics and Testing of campaigns
16
SECTION 3 How Can I Find Out More?
17
3a. Links
17
3b. Books
18
3c. Glossary of terms
18
3d. Abbreviations
20
3e. Help
20
1
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
Preface
The growth of the world wide web has been phenomenal. According to the Interactive Media
in Retail Group (IMRG), the leading industry body for global e-retailing, some £4.8bn was
spent in the UK online in September 2010, and £6.4 billion was expected to be spent online
at Christmas.
Government figures suggest that 7 out of every 10 businesses in the UK have a website
and many invest significant sums in designing, developing and maintaining their sites. But
how do you evaluate the effectiveness of your website? How do you improve its design to
maximize sales? The answer lies in Web Analytics and a whole new industry has developed
to measure and analyse internet data bringing with it a new set of tools and jargon.
This Guide provides an introduction to Web Analytics. It explains the jargon, looks at the
opportunities for marketers these tools offer and provides guidance on how and where to
find out more. Whether you plan to undertake web analytics yourself or employ a specialist
this Guide will help you understand what can be done and how to make the best use of the
information available.
2
Introduction
The Internet is a series of “interconnected networks” and can be traced back to the 1960s.
The World Wide Web is more recent and can be traced back to 1990 when Tim BernersLee conceived the usefulness of connecting documents on the Internet. These documents
needed to be linked by preparing each with hidden “hypertext codes”.
With the arrival of the World Wide Web we saw the birth of web pages, which collectively
became web sites. A massive industry therefore emerged; web site design and creation was
big business. Every company, whether large or small, wanted a presence on the World Wide
Web and the low cost made this possible. However early web sites were nothing more than
brochures, hastily pushed onto cyberspace. Gradually expertise grew and design was more
elaborate and well-considered. A new set of conventions and vocabulary emerged: linking,
keywords, stickiness, FAQs, Search Engine Optimisation (SEO), Pay Per Click (PPC).
The web site industry flowered surely. What is more fascinating is that silently an industry
grew in its shadow, an industry designed to monitor the performance of these web sites.
That industry has become known as Web Analytics. A professional body was created: The
Web Analytics Association (WAA). This body holds conferences, has created standards for
measures and has an ambitious education programme. The Official WAA Definition of Web
Analytics is: “the measurement, collection, analysis and reporting of Internet data for the
purposes of understanding and optimizing Web usage.”
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
The Web Analytics industry has more jobs than available candidates: it is one of the few
sectors still hiring! Here is a typical job description, this instantly gives an indication of the
type of work and tasks involved.
Web Analyst - London
, c£
50K
The Role
An outstanding individu
al with a real passion for
analytics. Can translate
disparate data systems.
complex information from
Must communicate sim
ple insights to a variety
a non-technical manner.
of
internal stakeholders in
This self-starter must be
able to work well with all
hierarchy, possess exc
levels of the business
ellent presentation skills
and be an enthusiastic
Not just a numbers per
and persuasive commun
son but an individual who
icator. can be focused on the
working to tight deadlin
customer experience whi
es and managing interna
lst
l stakeholder expectation
s. Key Responsibilities
zz To drive website opti
misation recommendatio
ns through actionable insi
zz To manage KPI das
ghts based on data ana
hboard production (daily,
lysis. weekly, monthly) and dev
business stakeholder
elopment with key interna
s on an ongoing basis.
l
This
incl
udes regular communicat
metrics to the approp
ion on key business
riate audiences in order
to keep them informed
zz To assist with the plan
and involved. ning and execution of ana
lytics and optimisation
audit the web tracking
initiatives and routinely
data to provide recomm
endations and best pra
web tagging. ctice for data collection
and
zz To present analysis
and findings on the key
customer journeys to inte
staff. rnal stakeholders and sen
ior
zz To assist in vendor
management periodicall
y by monitoring usage
recommendations. fees and new product
Pre-requisite Knowled
ge, Skills and
Experience zz Highly numerate with
extensive use of more
than one web analytics
zz Proven track record
software package.
of improving website con
version through analysi
zz Basic understanding
s and iterative investigatio
of the company’s busines
n. s model and peer-to-pe
zz Understanding of data
er proposition. warehouse fundamental
s
com
bine
d
with
analysis tools. business intelligence soft
ware
zz Marketing campaig
n performance evaluation
and
exp
osu
re
to
(paid search, SEO,
Channel promotions
affiliates, etc) crucial. zz Web technology fund
amentals along with the
evolving applications usin
g in page design. Desirable technical ski
lls zz Intermediate level of
understanding HTML zz Exposure to SASS/S
PSS software zz A/B testing experie
nce
This guide aims to break through the jargon and show marketers the range of opportunities
available to them when using Web Analytic tools and services. There are two options: first to
buy the services from established buyers, second to do web analytics in house. This guide
will identify the vendors offering services and it will also give introductory indications on
3
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
how web analytics can be carried out in house. These guidelines should help to build and
support brands. Web analytics emphasises the potential of the website as a promotional
tool and its use in business development.
Section 1 - What Are Web Analytics?
1a. The Web Analytics Industry
Let us start with a little history lesson and refer to Table 1. We start in 1990 when the World
Wide Web emerged. Early web sites had a counter at the top or bottom of each page. This
often boasted something along these lines: “we have had 10,000 visitors!”
Table 1: A Brief History of Web Analytics
Period
Event
Description
1990
Web counters easily available and
visible
Device which counts the number of page hits or visitors, this is
usually displayed on the page for all to see.
1995
Log File Analysis programmes
created
Analog, an early log file analysis program, was written.
WebTrends 1.0 took log file analysis further.
2000
Key vendors based on log
file analysis were firmly
established: Accrue, WebTrends,
WebSideStory, Coremetrics.
Problems with log files were becoming clear...Page caches by
ISP, Search robots.
20002006
JavaScript Tags arrive. Four big
vendors: Coremetrics, Omniture,
WebTrends, WebSideStory and
many mid-market names: Unica,
indexTools, ClickTracks etc.
JavaScript Tags emerged as a new standard for collecting
data from websites. A few lines of code are added to each
page. When a page loads, these tags cause data to be
sent to a data collection server. So work shifted from the IT
Department to a web analysis software vendor.
2006
Google Analytics offered free
In 2005 Google purchased Urchin and after some
development made the decision to offer the product free of
charge. This left many of the established vendors without
customers as anyone could perform web analytics easily and
at no cost.
2008
Yahoo! Web Analytics born
Yahoo! acquired IndexTools and immediately offered the
product free of charge but only to those paying to advertise on
Yahoo! Networks.
4
The early web counter can still be found on some pages, notably it is a badge of honour for
You Tube who show the number of hits of the most popular videos. On corporate websites
the counter is less common, as more complex information can now be derived from visitor
activity.
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
Behind every web site there is a computer server which was normally located in the IT
department at the company’s offices but now may be in the cloud. This piece of hardware
“serves” any requested information to the computer user. Inside every server there are log
files which are a record of all web page activity. In 1995 a log file analysis program was
written, this was called Analog and it took the position as the first major contribution to web
analytics, not least because it is free.
The first burst of growth in the web analytics industry took place between 1995 and 2000
where log file analysis was central to the operations of Accrue, WebTrends, WebSideStory,
and Coremetrics. The limitations of using log files were becoming clear as the numbers of
visitors was either overestimated or underestimated, this was due to such things as page
caches carried out by Internet Service Providers (to avoid pressure on networks), by the
activity of robots, and other issues.
A second method called Page Tagging (or JavaScript Tags) emerged as a new standard
for collecting data from websites. A few lines of code are added to each page. When a
page loads, these tags cause data to be sent to a data collection server. So work shifted
from the IT department to a web analysis software vendor. As a result Accrue, WebTrends,
WebSideStory, and Coremetrics became stronger as they could complement their services.
We also saw the arrival of smaller companies: Unica, indexTools and ClickTracks to name
just three.
It is worth noting that a series of mergers, takeovers and name changes has transformed
these early pioneers:
zz IBM purchased SPSS (2009) Coremetrics (2010) and Unica (2010)
zz WebSideStory rebranded itself as Visual Sciences (2007), Visual Sciences was
acquired by Omniture (2007) zz Adobe Systems purchased Omniture (2009)
zz Accrue was bought by Datanautics (2003) now Datanautix
zz Microsoft bought DeepMetrix (2006) and develop Gatineau, change the name to
adCenter Analytics then abandon the project (2009)
zz Urchin was bought by Google (2005)
zz IndexTools was bought by Yahoo (2008)
zz ClickTracks becomes part of Lyris (2006)
zz Comscore purchased Nedstat (2010)
There are three major search engines with advertising linked. These are Google, Microsoft
(with Bing) and Yahoo. It is important to look at their involvement in the Web Analytics
industry.
5
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
First Google purchased Urchin in 2005 and after some development made the decision to
offer part of the product free of charge in the next year under the name Google Analytics.
This left many of the established vendors without customers as anyone could perform web
analytics easily and at no cost.
Second, Microsoft bought DeepMetrix in 2006 and developed Gatineau which changed its
name to adCenter Analytics. Then, surprisingly, Microsoft abandoned the project in 2009.
There are indications in 2010 the idea may have re-emerged as “Campaign Analytics” as
part of the Microsoft adCenter initiative.
Third, Yahoo! in 2008 made a similar move to both Google and Microsoft in that it acquired
an established system, in this case IndexTools; this product was immediately offered free of
charge, but only to those paying to advertise on Yahoo! Networks.
This summary is intended to guide anyone intending to use external services. The names
crop up frequently and it is likely that the industry will see further changes, so it is important
to be aware of developments.
1b. Log File Analytics
6
Advantages of Logfile analysis. There are several advantages of Logfile analysis. The
data is on the company’s own servers, which makes it readily available. Also the format of
the data is standard, so it can be manipulated more flexibly than if it was provided through
a vendor with their own format. The benefit of this comes in the long term because the
analysis of historical data can be carried out with whatever software is available. Some page
tagging solutions have been accused of leading to “vendor lock-in” where it is difficult to
switch from one provider to another without substantial work.
Disadvantages of Logfile analysis. Conversely, Logfile analysis is criticised for providing
inaccurate traffic figures. One reason for higher traffic readings is that pages are visited
by robots in addition to humans. Software is programmed to “trawl” the worldwide web
continuously, particularly to update search engine databases, and there are other reasons
for “spider activity” which include experimentation, competitive intelligence gathering and
so on. Logfile analysis software cannot always distinguish human from robot activity. A web
page may not display fully before the user moves on to another website, this is not a “view”
but it is still recorded; page tagging detects if the page is viewed properly.
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
Figure 1: Typical Topline Logfile Analysis
7
Source: AWStats demo at http://awstats.sourceforge.net/
demo at http://www.nltechno.com/awstats/awstats.pl?config=destailleur.fr
A list of Free Log Analysers can be found at
http://www.prospector.cz/Freeware/Web-development/Log-analyzers/
1c. Dashboards
Dashboards are an essential part of web analytics so let us take a few minutes to describe
what they are.
Dashboards are single computer screens containing visual displays of performance
information, usually from several sources. They are important in Web Analytics because
most services deliver information in the form of a dashboard, and this can be customised to
meet your needs.
The dashboard is personalised for one person but several versions can be created from
the same data-set because several users may have different needs. The Web Analytics
dashboard is generally in real-time, with a delay of seconds; non-real-time displays may
have a delay of 24 hours or 48 hours depending on the information sources and service
used.
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
The term “dashboard” was taken from the console found in every motor car. Here the
marketing manager is in the “driving seat” and should be able to see his situation at a
glance. The marketing manager is expected to choose Key Performance Indicators (KPIs)
that help understand whether objectives are being met.
There are several steps necessary for effective dashboard design:
1
Selecting correct metrics
2
Selecting correct visual displays
3
Arranging the metrics in a meaningful way
4
Ensuring that important details are emphasized
5
Installing “alerts” to indicate changes that may require attention
Figure 2 shows the standard Google Analytics dashboard which will be familiar to users
of this service. Note that a line graph indicates fluctuations in traffic over the last month,
important numbers show the number of visits, the number of page views, the number of
pages viewed per visit. There is a world map indicating that most visitors come from the
USA. We can scroll down and see more information. There are numerous opportunities to
personalise the displays offered.
8
With the many thousands of possible metrics available to the digital marketer it is important
they focus on the measures and indicators from which they will be able to make strategic
decisions and changes to their current plans. With this in mind a good starting point by
which to decide the most relevant metrics to include in a dashboard is the Digital Marketing
Plan. There are many frameworks available upon which to base a plan and a popular
example is SOSTAC
Situation – where are we now?
Objectives – where do we want to be?
Strategy – what will it look like when we get there?
Tactics – what do we have to do to get there?
Actions – who will do what, when and how?
Control – What measures will we include in our dashboard?
Such a framework should work as a continual feedback loop with the outcomes from
measurement and analysis allowing the marketer to improve the situation by tuning their
tactics and actions.
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
Figure 2: The Google Analytics Dashboard
9
1d. Important Services Today
Because tagging is visible to the whole world we can go to any web site, inspect the code
and discover which service is being used. The Vendor Discovery Tool by Web Analytics
Demystified was written to automate the process. You can simply enter a web address into
the software (perhaps one of your competitors) and the “analytics vendor” is identified.
This is a free and transparent service so we can see the last 100 site enquiries made. This
is useful to see the importance of the different services. From this we also know that at
least 10 out of 100 companies use two or more analytic services, this is a minimum figure
because log file analysis use is not detected by the Discovery Tool. It is clear that “free”
Google Analytics dominates but that the paid-for Omniture has a strong hold on the market
place.
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
Figure 3: 100 Queries at the Vendor Discovery Tool in October 2010
(http://www.webanalyticsdemystified.com/vendor_discovery_tool.asp)
10
Section 2 - What Are The Opportunities For Marketers?
2a. Setting up your Web Pages for Analytics
For the purposes of this document we will examine the use of Java Tracking Code rather
than log file analysis. Log File analysis requires a greater technical expertise, for example
the page tagging service manages the process of assigning cookies to visitors; with logfile
analysis, the server has to be configured to do this. For these reasons we will focus on page
tagging. Furthermore we will use Google Analytics because we have seen that it is the
most popular method.
There are several steps involved in this:
1 Visit GOOGLE ANALYTICS at www.google.com/analytics
2 Click SIGN UP NOW (you will need a Gmail account)
3 Follow the instructions and seek any TUTORIAL HELP during your journey
There will come a point where you are instructed to copy some text and to insert this into
the HTML of your web page. Java tracking code will work no matter where it’s placed on
the page. The reason it’s recommended that JavaScript is placed at the end of the page is
because javascript blocks the browser from rendering the page while it’s executing, but if
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
it’s at the bottom of the page (right above </body>) then the page will have already (mostly)
rendered, thus improving the “apparent” load time of the page.
After several hours the ready-made dashboard will start to fill with numbers, visual displays
will come to life. This dashboard contains metrics that you are likely to use but it can easily
be changed to accommodate your needs.
On the dashboard prepared earlier, with extracts shown here we see, for the past month:
zz Number of visits as a line graph
zz 1 and a half minutes spent on the site
zz Site usage: 1163 visits, 2517 page views
zz Visitors: just over a thousand
zz Site usage: 2 pages were seen per visit
11
zz Traffic sources: 11% came direct, 75%
zz We can see the keywords used in those
came via search engines and the rest
were probably links on other sites
zz We can even see which pages were
zz We can see which search engines were
used (Google dominates)
search engines
seen the most (and least)
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
2b. Bounce Metrics
For many years we have heard the term “stickiness” used in relation to web sites. If a web
site is sticky then we attract a visitor and keep them looking at our site, hopefully they will
take some action such as requesting further information or placing an order. Web Analytics
has a word to describe the opposite of stickiness – it is the word “bounce”. Bounce occurs
when a visitor arrives at a web page then leaves the site, effectively a single visit.
The dashboard extract below shows a bounce rate of 67%, it means that 67% of visitors
took one look at my page and decided it wasn’t for them. Only a third of visitors decided to
look further into my site.
This can be interpreted in many ways. One interpretation is that I have been promoting my
website to the wrong people, I may have paid for the wrong “keywords” or I may have given
misleading descriptions in my pages or in the hidden messages I write for search engines.
12
Another interpretation is that my “entry pages” are badly designed; I need to modify them to
create something far more interesting.
Avinash Kaushik, an authority on bounce rates, said “It is really hard to get a bounce rate
under 20%, anything over 35% is cause for concern, 50% (above) is worrying.”
So we can see I have a real problem that needs to fixed, and a simple number has allowed
us to see the problem in the first place.
It must be said that specific pages, such as blogs and case studies, especially when used
as a landing page within a customer retention campaign, see high bounce rates because
the visitor is already very familiar with the content on other pages on the site or is just
dipping in for breaking news. It’s therefore important to say that whilst bounce rate is a
key performance indicator in website analytics it should be used in conjunction with other
measures to give a well rounded view.
2c. Segmentation Filters
As marketers we are all taught about “Segmentation, Targeting and Positioning”. We must
divide the population into similar groups, only then can we decide what targets to select.
Once we have a target or several targets, we can then position our product in the mind of
the prospect. Web Analysts have taken the aspect of “segmentation” (although targeting
and positioning are not mentioned in those words).
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
Table 2: Segmentation Bases Available or Unavailable through Web Analytics
Can be created with a form or
adding service
Base
Standard
Demographic
Language preference
Age
Sex
Education
TEA
Occupation
Social class
Geographic
Country
Geographic region
City
Timezone
Postcode area
Zip code area
Behavioural
Online purchases
Page visits
Bounces
Conversion behaviour
Previous web site
Referring website
Browsing behaviour
Keywords used
Pages chosen
Search engine used
Hour of the day
Mobile phone user
Connection speed
Browser used
Benefit sought
Loyalty
13
Psychographic
Lifestyle
Personality
Geo-demographic
ACORN
MOSAIC
PRIZM
Dashboards have a set of pre-defined segments and the opportunity to build more. In Table
2 we use the traditional bases of segmentation from marketing to list some of the choices
on offer by web analysts. Where segments are NOT available they can be created by
presenting web site visitors with a form and asking them to select an option such as job title
or hobby. Other non-standard segments can be created by purchasing and fusing data, as
is the case with geo-demographic systems or some vendors have proprietary systems that
extend segments.
Adding segments to the dashboard is as easy as dragging or clicking as we see in Figures
4 and 5 (see next page). Figures 6-9 shows how segmentation by mobile phone use and
then geography appear on the visual display. It is interesting to see that USA visitors have
the average bounce rate (68%) but visitors from India are down at 50%. Here we see the
value of applying the segmentation filter – it shows that Indians are more interested in the
pages than Americans.
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
Figure 4: How to Create Extra Segments (Yahoo Analytics)
Figure 5: How to Create Extra Segments (Google Analytics)
14
Figure 6: Dashboard Showing Mobile Phone User Segments
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
Figure 7: Segmentation of Visitors by Country (map)
Figure 8: Segmentation of Visitors by Country (table)
15
Figure 9 Dashboard showing USA/India visitor Bounce rate differences
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
2d. Conversion Metrics and Testing Campaigns
It is important to remember that all web sites have a purpose, and usually it is to persuade
the visitor to carry out some action. It might be:
zz To read an article
zz To sign up to a newsletter
zz To request a brochure
zz To make contact
zz To place an order
When this objective is achieved, then we can call it a conversion. Yes, the term “sales
conversion” springs to mind, but the web analytics use of the term can be stretched to other
actions. One display option on most dashboards is the funnel. Figure 10 is a sales funnel
used by the vendor Unica. If we follow it from the top we see that 2069 people entered an
offer code, 473 started their shopping cart. There were some who abandoned their cart
because 470 began the checkout but only 399 completed an order form.
We can substitute different actions at each point and instantly see the dropout at different
points and “fix” the problem. In this case there was something after the offer code that
16
caused over 75% of people to go away.
Figure 10: A Sales Funnel used by the vendor Unica
Finally it is important to mention A/B Testing. This is very simple, and direct mail marketers
will be familiar with the idea. If we have two executions of a web page. Perhaps two different
offers – let’s say one of the campaigns is centered on low price but the other on quality.
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
We simply call the first “A” and the second “B”. These two pages are created and the
server is told to alternate the viewing: so A goes to the first visitor, B to the second, A to the
third and so on. Then a conversion funnel is created for both options and we can see the
difference. This is simple but effective experimentation. Another test you may come across
is MVA which means Multi-Variate Analysis, it goes beyond testing two versions (bi-variate
analysis) and tests many aspects. Clearly this is a specific study area with many guidelines.
This guide has shown you the opportunities available to you when using Web Analytic
tools and services. Clearly individual needs differ but this should be a starting point for you
whether you want to buy vendor services, do it yourself or want to build a career in this new
sector.
Section 3 - How Can I Find Out More?
3a. Links
Marketers should consider the following links to more detailed information references in this
guide.
The Web Analytics Association (WAA)
http://www.webanalyticsassociation.org/
A Proposed Code of Ethics
http://waablog.webanalyticsassociation.com/2010/09/web-analytics-code-of-ethics.
html
Free Visitor Counters to add to your web pages
http://www.website-hit-counters.com/
http://www.free-counters.co.uk/
Web Analytics Demystified: free cases, whitepapers and even books!
http://www.webanalyticsdemystified.com/content/index.asp
Vendor Discovery Tool
http://www.webanalyticsdemystified.com/vendor_discovery_tool.asp
Web site health check up, free. Try it now
www.bartlettinteractive.com/evaluator Google Analytics Web site
http://www.google.com/analytics/
Yahoo Analytics Web site
http://web.analytics.yahoo.com/
(Google) Insights for Search
http://www.google.com/insights/search/#
17
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
Google Trends
http://www.google.com/trends
Microsoft Analytics links
https://adcenter.microsoft.com/
http://en.wikipedia.org/wiki/Microsoft_adCenter
http://advertising.microsoft.com/uk/learning-centre/search-advertising/adcenterwhats-new
3b Books
Arikan, A. (2008) Multichannel Marketing. Metrics and Methods for On and Offline Success,
Sybex
Bradley, N. (2010) Marketing Research: Tools and Techniques, Oxford University Press,
Oxford
Clifton, B. (2008) Advanced Web Metrics with Google Analytics, Sybex, Wiley
Kaushik, A. (2007) Web Analytics: An Hour a Day Sybex, Wiley
Kaushik, A. (2010) Web Analytics 2.0 Sybex, Wiley
18
Mortensen, D. R. (2009) Yahoo! Web Analytics. Sybex, Wiley, Indianaplois, Indiana
Peterson E.T. (2004) Web Analytics Demystified: A Marketer’s Guide to Understanding How
Your Web Site Affects Your Business. Celilo Group Media Peterson E.T. (2005) Web Site Measurement Hacks. O Reilly ebook
Sterne, J. (2010) Social Media Metrics. How to measure and optimise your marketing
investment, Wiley, New Jersey
3c Glossary of Terms
These definitions were created by Nigel Bradley. You are also advised to consult
terms created by the Web Analytics Association which act as an industry standard.
Bounce Occurs when a visitor arrives at a web page then leaves the site, effectively a
single visit.
Click A positive and deliberate action made by a computer user, usually by depressing the
left button of a mouse.
Clickthrough rate (CTR) A measure expressed as a percentage. This is the number of
clickthroughs divided by the total number of times the page in question has been served
(the number of page impressions).
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
Dashboards Single computer screens containing visual displays of performance
information, usually from several sources, a type of portal.
E-metric Performance indicators derived from traditional research or web analytics.
Log file A record of the download activity of a specific set of web pages. Log file analysers
are used to make sense of the data. Useful to create online measures for marketing.
Stickiness A phrase used to describe whether a site is able to hold a visitor. This can be
quantified by the length of visit recorded in log files.
Tagging Observing the movements and behaviour of people using cookies, RFID or other
techniques, this may be online or offline.
Usability tests Term used to describe the procedures used to discover whether hardware
or software is easy to operate.
Web analytics Also known as website analytics. The capture and processing of data from
software and hardware that produces performance indications (e-metrics).
Web servers Computers used to store web sites and their corresponding databases.
19
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
3d Abbreviations
CPC
Cost per click
CTR
Clickthrough rate
HTML
Hypertext Markup Language
http
HyperText Transfer Protocol
IP Internet protocol
20
KEI
Keyword Effectiveness Index
KPI
Key Performance Indicator
PPC
Pay Per Click
SEM
Search Engine Marketing
SEO
Search engine optimisation
URL
Uniform Resource Locator
URL
Universal Resource Locator
WWW
World Wide Web
3e. Help
Tutors at Cambridge Marketing College blog on this and other research topics at
http://tutors.marketingcollege.com
and through the College’s Digital Marketing website at
http://www.digitalmarketingprogramme.com.
Cambridge Marketing College offers 2 digital marketing qualifications: the CIM CAM
Diploma in Digital Marketing and the CIM CAM Diploma in Managing Digital Media.
More information on these courses is available at:
http://www.marketingcollege.com/default.asp?edit_id=1345-28.
CAMBRIDGE MARKETING COLLEGEGUIDE TO WEB ANALYTICS
About the Author
This Guide was written by Nigel Bradley who is a Senior Lecturer in Marketing at the
University of Westminster. With a Masters degree in Product Management and Marketing
from Cranfield, Nigel spent his career in marketing research agencies before joining
academia in 1996. Nigel’s book Marketing Research: Tools and Techniques was published
by Oxford University Press. He has been a Tutor with Cambridge Marketing College since
2003.
21
C A M B R I D G E
S T
J O H N ’ S
I N N O V A T I O N
M A R K E T I N G
C E N T R E ,
C O W L E Y
I N F O @ M A R K E T I N G C O L L E G E . C O M
•
R O A D ,
C O L L E G E S
C A M B R I D G E ,
C B 4
0 W S
W W W . M A R K E T I N G C O L L E G E . C O M