Abstract: This paper describes the components that constitute
a contemporary web analytics application and discusses the
impact and the implications of the technology paradigm shift
towards rich Internet applications (RIAs) on web analytics
applications in Web 2.0. The performance and behavioral
measuring mechanisms collect data from various source
points. These sources depend heavily on the implementation
characteristics of each web application. For example, while
the amount of detail data registered in the access log file
(ALF) by the web server is sufficient for any non RIA
application, it is heavily reduced for any RIA, since the model
of interaction between client and server is heavily shifted
towards the client [1]. In RIAs most processing and
preparation of data is performed locally, at the client site by
the browser, so the client handles all key clicks, as well as the
main interaction with the user interface of the application.
Communication with the server mainly covers transfers and
requests of data to and from the database of the web
application. The two basic technologies used for analytics,
tagging and log file analysis are described and discussed. A
hybrid solution that addresses the issue of the reduced ALF is
proposed. Additionally extraction techniques for specific
crucial data needed in order to scrutinize the performance
operationally, as well as from the business perspective of the
web application enriches the analyzer's database and provides
measurements and data using terms and descriptions familiar
to the operational personnel.