Analysis tab

The following options determine how Deep Log Analyzer identifies unique visitors. DLA applies heuristic algorithm to distinguish different visitors by IP address or host name, user agent and other values. Identifying visitors is important as it allows to correctly determine such values as user loyalty and visitors return rate. Accuracy of most reports depends on how correctly unique visitors are identified.

DeepTracker method is the most accurate tracking method available in web analytics today. This method uses javascript to set permanent user ID cookie on visitors computer. When the user with this cookie returns to the website later it's precisely identified as a returning visitor regardless of his IP address.

This method uses session cookies for sessions tracking. This allows Deep Log Analyzer to identify series of different page views and hits as a visit

DeepTracker Code inserted in web pages collects extended client information about visitors that is shown in these reports:

This method requires DeepTracker Code inserted in all pages of the website. DeepTracker code is a small piece of HTML code. If your website design uses includes for displaying header on each page, it is the best place to insert the DeepTracker Code. DeepTracker code should be inserted right below <body> tag, although other positions will work too.

Put dla_tracker.js and dla_tracker.gif files in the root folder of your website. These files can be found in "{Program Files}Deep Log Analyzer\DeepTracker\" folder. If your website has more than one directory with web pages, these 2 files should be uploaded to those folders as well in case DeepTracker Code was inserted in web pages there.

3.	Cookie field must be present in log files. You should set up log format on the web server to include Cookie filed. This is done differently on MS IIS and Apache. Follow these instructions to set up cookie field in your web server.

4.	If any of above requirements is not met Deep Log Analyzer will use IP Address+User Agent tracking method.

This method uses a combination of visitor's IP address and User Agent (OS +Browser) to identify visitors and sessions. It does not require any changes in website design, although it is less accurate. IP addresses can change over time even during single visit. This fact inflates number of unique visitors and visits the software reports.

•	Use User Names rather than IP Addresses...: check to identify visitors by user name instead of an IP Address.

Note: This option allows DLA to identify visitors very precisely, but only applies when users log in to your website. This works only on secure sites that require user authentication. Only users authenticated by methods that record the user names in web server log file will be recognized, for example is Apache server '.htaccess' method. If there is no value in this filed DLA still applies its standard user identification algorithm.

•	Visit Timeout: type the site's visit timeout value. Visit timeout is the amount of idle time, i.e., when the user has not clicked any links on your site, the user's visit is considered finished.

Note: You only need to enter a value in the Time Adjust field if your web server is located in a different time zone and if the exact time of the hits is important to you.

•	Date Format: choose a date format. Options are Auto, MM/DD/YY, DD/MM/YY, and YY/MM/DD.

Note: If you choose Auto, Deep Log Analyzer will try to determine the date format native to the log file, as log files may differ between web servers. Take into account that it's not always possible to correctly determine the log file's date format. For example, DLA can not automatically determine date format of 01/06/04. If you see any inconsistent dates in the reports you should view the log file and define date format here.

- If you change any of the settings on this page for the existing project you have to reimport log files in order the changes to take effect.