Simple Split Tests

Simple split tests are simply on or off. These split tests simply compare one campaign to the control (what the customer sees normally). In order to do this, you must select which campaign you want to perform a split test on and press the start test button next to it. This will present the screen below:

From here you can what percentage of people will see the alternate variant of the page as opposed to the existing standard. For example, tin the above screenshot 50% of the people will see the new popup from this campaign while the other 50% will not see any popup at all. Once you have decided on a percentage, you should press Start Test to confirm this test to begin. 

As you can see from the screenshot above, the module has been placed into a simple split test. 

Using Google Analytics

NOTE: This section is fairly technical. If you're having trouble, please contact your tech team, or whoever is responsible for Google Analytics on your website.

If you wish to integrate with Google Analytics for the split test, you must first have it set up on your site. Additionally, Universal Analytics compatibility can be enabled on a per account basis. This process is beyond the scope of this article but is given in detail  on the google website. You must also make sure your site is correctly tagged as per our Tagging your website article.

To activate Google Analytics on a simple split test, check the box and you will be given the following options:

Here you can select a slot number (Dimension Index), a GA variable name, GA Category and GA event name. Our system will send these to your google analytics whenever a user interacts with the test using the following code:


_gaq.push(["_setCustomVar", slotNumber, gaVarName, "True", 2]);
_gaq.push(['_trackEvent', 'Intilery', gaEventName, '', 0, true]);


ga('set', 'dimension' + gaSlotNumber, gaVarName); 
ga('send', 'event', gaCategoryName, gaEventName, {nonInteraction: true});

where gaVarName will be a string of the form:

[GA Variable Name]-[control] or [GA Variable Name]-[variant]

and gaEventName will be a string of the form:

[GA Event Name]-[control] or [GA Variable Name]-[variant]

E.g. If you set a GA Variable Name: "aVariable", and the customer hits the test variation, the gaVarName will be "aVariable-variation", while if they do not hit the test variation the gaVarName would be "aVariable-control". The same applies to gaEventName. In this way your Google Analytics can analyse traffic through your split test.

Still need help? Contact Us Contact Us