How do I set up an A/B Test?

The following article highlights some KPI metrics that can be measured and how the test should be set up to attain valid results.

What KPI’s to measure with an A/B test?

We Recommend on of the two options below:

  • Conversion Rates are an indicator for the increased purchase probability due to the usage of the Size Advisor - which leads to a decreased insecurity about sizing of visitors.
  • The Average Revenue per Visitor incorporates effects on changed spending behaviour (e.g. more products or more expensive products are bought). Typically the Average Revenue per Visitor gives a more comprehensive picture as this KPI measures how much more revenue is generated per customer through use of the Size Advisor.

How do I set up the test to measure these KPIs?

In both cases, these KPI’s need to be comparing visitors on the A-side who click the default size chart vs visitors on the B-side who click the ShoeSize.Me Size Advisor. The test should not be measuring against and comparing all visitors. 

 

AB test1

Implementation

 

AB Test 2

IMPORTANT: Never show 2 different sources of size help directly on your product detail page. If you provide any other size information (e.g. Size charts, brand comparison charts etc.) to your customers, do NOT show this on the Variation Side (B Side) as multiple sources of size information confuses other online shoppers and negatively impact conversion rates.


Need help? Contact us at support@shoesize.me