Google User Defined tracking, or “custom variables” are name-value pair tags that you can insert in your tracking code in order to refine Google Analytics tracking. With custom variables, you can define additional segments to apply to your visitors other than the ones already provided by Analytics.
The implementation is for legacy tracking for ga.js. New tag is using customer dimensions as a way to implment custom tracking to help marketers do segmenetation
The Google Analytics’ user defined report allows an analyst to compare visitors from segments that you have defined. I will go through several types of segmentations that you could possibly set. You can define these segments by calling a line of code/function in your web page. So every time a page with the code is requested, a custom value is captured and stored in the user defined variable.
The main approach to executing this is to simply apply a code like this:
As an example, one of the sites I work on has English and Japanese sections. Each section will have a value to identify its section, defined as “/viewed/English” or “/viewed/Japanese”.
Once you have these segmentations in place, you’ll be able to see how users are behaving differently in each section.
Visitor Type Segmentation
This is a powerful segmentation to get an actionable insight to who is behaving differently. Let’s say you have a form entry that leads to a confirmation page. Assuming the form has a field with values “Engineer”, “Project Manager” and “Director”, and when the form is completed, one these values will be stored into user defined variable.
At a certain point, you may learn that Directors are likely to complete the form and convert. This will tell you something about your visitor that you didn’t know.
Landing Page Segmentation
You may have custom landing pages to serve different campaigns or promotions. Identifying the landing pages through user defined report would be a powerful method to analyze effective landing pages or even its campaign effectiveness.
(Example 1) Say you have two different direct mails with a different slogan or description. You can have two different friendly URLs set in each of the direct mail. Direct mail A would have a landing page A, and landing page B for direct mail B. When traffic and performance for landing page B performed better than landing page A, that could mean that the direct mail B’s strategy was more effective than version A.
(Example 2) You can have one campaign with a landing page, but a page can receive traffic from other various sources. Setting two different landing pages and segmenting it through “user defined” variables can show you which landing page is more effective while testing various traffic sources.
Above examples can speak as an example for A/B test. Another example to perform A/B test is testing different call to actions (CTA). Say you have two links; one with an image link and other as a simple text link.
You may distinguish these different CTAs (or possibly link type, position, etc.), and store it into user defined report to assess which criteria performed better.
Google analytics has a sophisticated campaign tracking method. However, you can choose to use user defined report to parse certain attribute within URLs and apply it to the report. One possible example of such application would be existing links with an identifier in the URL (not compliant to Google’s campaign tracking), where the links are located at two different sites.
Site abc.com with a link “yoursite.com?source=123”
Site xyz.com with a link “yoursite.com?source=456”
Your landing page could parse these source codes and allocate proper value to a segment. Therefore you should be able to assess the referrers’ performance and its effectiveness.