Due to the growth of
WWW related technologies, the number of web sites on the Internet has increased
rapidly, and human daily life is beginning to depend on such sites like
shopping sites, official sites of enterprises, promotion sites of events and so
on. Each website has different
types of information or content e.g. articles, blogs, newsletters, and training
videos.These web sites contain a variety of content and complex link
structures. Therefore it also requires an understanding of what content
draws users attention and how users interact with that content. As content on
any website is one of the most important element, we need to optimize
content. For content optimization we
need some metrics to tell us how each aspect of the content performs. How does
the content on the web site affect the traffic patterns? Does it lead users to
the site? Is there content on the site
that performs better than we expect it to?
Web site administrators, who are constantly required to improve the
easiness of use of such large and
complex web sites, need to analyse user’s needs and demands. Web Analytics
tools are most important way to gather and analyse this information. Web analytics is widely used in commercial settings for
making business decisions and improving the customer experience. (Jacoby and Luqi et al, 2007; Sen et al.,
2006; Srinivasan et al., 2004). Phippen et al. (2004) present a definition
of web analytics that has emerged through its application to e-commerce and
for-profit organizations. They cite Aberdeen group‘s definition of advanced web
analytics as a tool for monitoring and reporting of website usage so that
enterprises can better understand the complex interactions between web site
visitor actions and web site offers [and] leverage insight to optimize the site
for increased customer loyalty and sales. One of important source of
information and analysis is Web access log. A web access log is a
time-series record of users’ requests which are sent to a web server when a
user does some operation on a web page. Analysing the logs is very useful for
the administrators to understand users’ behaviour on the web site. For web
access log analysis, statistical methods, like Google Analytics, Yahoo
Analytics are widely used. The results of statistical analysis contain bounce
rate, page views, page browsing time, and so on. Google Analytic, a free service offered by Google, generates detailed statistics about the visitors to a website. The product is aimed at marketers as opposed to webmasters and technologists from which the
industry of web analytics originally grew. Google
Analytics provides us reports for Daily, monthly, yearly tracking of web visits
which are graphed over time. We can create custom reports organized in the way
we want to show metrics and dimensions like page views, bounces, visits, and
revenue for each source and keyword. It provides reports to track visitors by
identifying parameters like pages visited by user, how long they stay, entrance
pages, location, operating system, monitor resolution etc. But these reports are very lengthy and hence
difficult to analyze on a whole, mainly due to the nature of the data. One such
example is the keyword based report. ‘Keywords’ are the words entered by user
to search for a specific content or web page. ‘Keywords’ are entered by users
based on his requirement, understanding and perception. Different users will
mostly enter diverse keyword strings although all of these may point to the
same web page. Due to this inherent nature of the ‘Keywords’ the Keyword based
reports can be very lengthy. Most of the website administrators may not go
through entire report but only through the top 10 or maybe top 20 keywords to
judge the content. Although for some website this might be sufficient but for
large webpages this will not provide the complete picture. The instruments of study are based on the relevant
literature and the reports generated by Analytic-Tools like Google Analytics.
In this paper two metrics have been choose to analyze the Keyword based report.
These metrics are (i) Organic Keywords (ii) Bounce Rate for monitoring traffic.
These are the two top priority metrics that will help to identify the various
keywords through which website is getting traffic from the search engine.
Bounce rate is used to identify the visitors which bounced back from landing
page without navigating further on the website. To develop this tool extensive
analysis was done on the data set of www.slideworld.com. This website specializes in templates and presentation on
various topics. While conducting analysis on its Google analytic reports it was
identified that users typically land on the site by searching for a specific
presentation topic. The Keyword based report was found to include more than 100
keywords which lead the most visits on the site. Some generic keywords like ‘ppt’ and extremely generic like
‘of’,’on’,’what’ etc were discarded in order to identify the traffic being fetched by targeted keyword on
the website. We combine keywords that have relevance meaning then we get
a clearer picture of the whether content is useful or need to improve.
The main contribution of our work include -
•
Algorithm named KBWOT to measure Keyword
similarity and combine Keywords based on factor of similarity i.e. 80% similar
or 90% similar based on user preference.
•
Implemented KBWOT procedure using Perl scripting
language. The KBWOT procedure generates XML
•
Flex based UI tool to reorganize the Keyword
based report based on the ‘similarity factor’ and provide a holistic view of
the similar keywords.
•
Process to combine the metrics like the visits,
bounce rate for the similar keywords by using the concepts of weighted mean.
•
Process to compare & analyze Keywords with
suggested Keyword provided by the user.
Comments