UNLedu Framework 3.1 Tracking Changes

From Web Developer Network Wiki
Jump to: navigation, search

In November 2012, an issue was discovered with department-specific Google Analytics implementations may cause inaccuracies in your GA reporting. This issue appears to be related to specific circumstances coupled with the move to the UNLedu 3.1 framework. Sites in UNLcms were automatically corrected at the time; if your site was in UNLcms in November 2012 you do not need to take further action.

Sites which have implemented UNLedu framework version 3.1 and are outside of UNLcms will need to have a portion of the GA tracking script removed. If the site has not moved to the 3.1 template, please note you will need to make this change at the time of conversion.

For sites in the 3.1 template, please remove the following line from your sitewide Google Analytics tracking code ASAP:

_gaq.push(['_setAllowHash', false]);

This tracking issue appears to only affect sites which are subdomains of unl.edu (ex: yoursite.unl.edu). For sites which are directories on unl.edu (ex:www.unl.edu/yoursite), there don't appear to be any negative effects.


Compromised Metrics

Data from the time your site was moved to the 3.1 template and the time the above line is removed may have inaccuracies. The following metrics may be compromised.

Bounce Rate

Bounce rates may be reported as 100%. This is incorrect.

Unique Pageviews

Unique pageviews are reported to be the same as total pageviews. The total pageviews number is accurate, the unique pageviews is not.

Time on Page

Reporting as 00:00:00. This is inaccurate.

Exit Rate

Reporting as 100%. This is inaccurate.

Page Entrances

May be inflated.

New Visitors vs. Returning Visitors

The data may incorrectly skew towards more new visitors.

Un-compromised Metrics

Data related to total pageviews, goal tracking, user events, custom variables, page speed, search keywords and landing pages, campaign analysis, traffic sources, mobile users, visit date and time, and the majority of statistics have not been negatively affected.

Personal tools