Although the SAS Data Surveyor for Clickstream Data processes
standard Web server log files, these files are limited in the following
ways:
-
They provide a limited set of data.
-
The data is captured only from
the perspective of the Web server.
-
The data includes every request
to the Web server, even for files that are typically not of interest
(such as image requests and spider or robot requests). This situation
results in larger data volumes and a need to perform a great deal
of filtering of the files.
-
Some user actions are not captured.
For example, browsers commonly cache pages. In that case, the use
of the forward and back buttons in the browser does not result in
a new request to the Web server. This processing results in user activity
that is missed in the Web log.
These
limitations of standard Web logs can be overcome with the use of a
method of client-side (browser) data collection called page tagging.
The page tagging method does not rely solely on the information that
a Web server can gather. Instead, it uses the Web browser to gather
data not normally logged by the Web server. The browser can gather
this data because a small piece of code has been inserted into each
page for which data is desired. This piece of code is known as a
page tag.
The
page tag runs inside of the user’s Web browser when the user
accesses a tagged page. The tag code has access to additional information
from within the browser that is not normally available in a standard
Web log. Once this data has been accessed in the browser, it is collected
by sending it to a standard Web server. The Web server then stores
in its Web log file only the requests for those pages that were tagged.
When a Web server is used in this way (to collect clickstream data
from tagged pages), it is referred to as a
clickstream
collection server. For a list of the data collected
by the clickstream collection server, see
JavaScript Page Tag Code.
Working
together, the page tag code and one or more Web servers configured
as clickstream collection servers provide a framework for client-side
data collection. The actual data that is tracked is controlled with
the page tagging code that you insert. For more information, see the
SAS Data Surveyor for Clickstream Data 2.1 Page Tagging JavaScript
Reference at
http://support.sas.com/rnd/gendoc/clickstream21M1/en/.