|
Web
sites are visited by many different types of people and for many reasons.
Identifying particular characteristics of these people and the
reasons they access the web site can be helpful in meeting site objectives:
to inform and educate, and/or to make a sale or site conversion.
Characterizing the behavior of web site users can be helpful in guiding the
optimization of web site content and structure. For example, characterizing
web site behavior can help identify which pages should be linked together,
and it can help identify what content pages and conceptual links should be
added to improve website usability and more clearly define web site funnels
toward site goals.
Applying
behavioral classification algorithms to the analysis of web site log data
can provide two main pieces of information: (1) behavioral links (not
necessarily physical html links) between pages on the web site that users
of the site tend to follow most often, (2) subgroups of the general user
population called personas that characteristically access different content
on the web site.
This
page includes example analysis of real log file data from a web marketing
web site to illustrate the concepts of user personas and behavioral
links. Using the results of the
analysis, some suggestions for improving site performance are offered.
Among the suggested solutions is the application of eye-tracking analysis
to identify how qualitative aspects of the web page such as characteristics
of pictures and the words used in links affect how people interact with the
site and affect the behavioral links among pages.
|