Introduction
An A/B Test in an e-mail campaign consists of setting up several versions of the campaign and distributing them according to a defined distribution and criteria.
This article details the various parameters involved in setting up an A/B Test.
Description
A campaign with A/B Test has a minimum of two different versions, and a maximum of 5. These versions differ according to a criteria, here called the comparison criteria
Comparison criteria
A/B Testing enables you to keep the best results between mailings according to a single criteria at a time, from among :
- Email subject
- Email sender
- Email content
Please note that it is not possible to select several comparison criteria. The aim is to isolate and ensure that the test is not biased by several differentiating elements.
Types of configuration
Two types of configuration are possible for A/B Test campaigns:
-
Without generalization: 100% of the target is divided equally between each version.
For example, 50% of the sample receive version A and 50% version B. -
With generalization: a defined percentage of the target receives the winning version of the A/B tests, according to the chosen metric.
For example, 5% of the sample receive version A and 5% version B. The remaining 90% of the target will receive the best-performing version.
Conclusion
When an A/B Test is activated on a campaign, several settings are available, and additional tests are required before the campaign is broadcast.
Further reading
How to carry out an A/B Test without generalization in an email campaign?
How to carry out an A/B Test with generalization in an email campaign?