Rule variations

Website rules can be optimized and tested by creating variations of a rule and comparing their performance. For example you might want to check the effect of a popup offering free shipping vs a popup offering a discount coupon.
In order to do that you can create a rule with the two variations and allocate certain traffic to each in order to compare them.

Setting up rule variations

To set rule variations create the first variation and then click the add variation button.
It will create a new variation with the same action as the original one which you can change.

Control group

A control group is used to compare visitors exposed to the rule action against visitors who match the rule segment but are not shown the rule action. This allows to see the rule effect on conversion and other metrics.
A control group is created by default so activating the control group is as simple as diverting traffic to the control group using the traffic allocation bars in the optimization and testing box (see below).

Allocating traffic

Traffic allocation between variations and control group is done using the slide bars in the testing and optimization section. 

Only active variations are shown, use the slide bars to change traffic between them.

Automatic Optimization

Barilliance automatic optimization uses machine learning to optimize the performance of rule variations. To activate automatic optimization you need

  • To add at least one variation
  • To define the rule goal 
  • Check the automatic optimization checkbox.

Rule goals

Rule goal can be one of: Conversion rate, Revenue, Average order value, Time on site, Number of add to carts, Bounce rate and URL goal.

URL goal is a url (or part of url) which the user will step on after being exposed to the rule.

Viewing variations performance 

Variations (and control group) performance can be tracked in the Barilliance dashboard (launch Feb 2020) and using Google Analytics.

Firing google Analytics events

for more information on firing Google analytics events see [TBD] 


Did this answer your question?