As web servers record transactions in logfiles it makes them easily read by programs that can provide data the popularity of a site.   Just over a decade ago statistics given regarding a website consisted of not much more than how many hits were made to the web server.  As websites consisted on a single HTML file this was reasonable, yet as images to HTML files was introduced, this statistic became almost useless.

In view of this two units of measure were released that could gauge the amount of activity on a web server. These units were called “page views and visits”.  A web server defined page views as a request for a page whereas visits were defined as a sequence of requests.  These requests expired after a time of inactivity and originated from a uniquely identified source, or client.  Today they are still in use but are considered out of date by many.

Difficulty to identify unique clients came about as a result of search engine spiders, web proxies and ISP’s.  In response, log analyzers began tracking visits through cookies and ignored requests that came in from spiders.  However, a problem presented itself to logfile analysis in the form of web caches.  No response would be received by the web server if a repeat visit was retrieved from the browsers cache.   The result is that the repeat visitor’s movements on site would be lost.  While caching can be overcome by configuring web servers, this results in a less than optimum experience for human visitors.

Financial implications of Logfile Analysis

Logfile can only be performed in-house which entails a one-off software purchase.  There are however many open source tools available that cost nothing.  If this is not an option, many proprietary vendors can offer a maximum number of page views per year yet this does entail additional costs.

You need to store and archive you own data with Logfile analysis which is apt to grow larger very fast.  The cost implication for IT overheads can be expensive and if you start running out of storage space this may present serious security issues.  There is significant maintenance of software and security patches required with Logfile analysis.

Whether Logfile analysis is the route for you depends on technical expertise available and the depth of information required and the amount of data that may need storing.  Logfile analysis also involves the cost of interpreting and analysing visitor behaviour on site.  This entails turning raw data into useable information.  Either this can be done by hiring a third party professional, hiring a web analyst in-house or training someone to do the work.

Logfile analysis falls under the heading “on-site” web analytics. It entails the collection of data in order to process it for reporting or research purposes. In addition to Logfile analysis there are other data sources that are able to augment the data collected.  These include e-mail response rates, information regarding sales or leads, data derived from direct mailing campaigns as well a data based on user performance.