What is an A/B test?
In marketing, A/B testing is a decision-making tool that allows you to choose between 2 versions of the same campaign, in which a single element is changed (the subject, the visual, the target...). The goal is to determine which version is the most efficient.
In the context of Predictive CDP, an A/B test allows you to compare two targeting methods. This test will allow you to statistically determine which targets generate the best performance between those identified by Predictive CDP and those generated via another method of your choice.
How to set up an A/B test?
Here is a step-by-step guide to help you set up an A/B test:
👉 Use your own targeting method to create a target that can be based on one or two of your criteria, for example. You will then obtain volume A.
👉 Create a campaign in the Predictive CDP as you usually do.
⚠️ The audience volume of your predictive campaign must be greater than the volume obtained by your targeting method. The Splio teams will equalize the volumes during the analysis so that they are comparable. If necessary, once you've created your predictive campaign, you can adjust the audience volume from the Audience section of the Campaign details side panel.
⚠️ Some advice
While we recommend you build your two targets in the order shown above, you can still choose to start by creating your target in the Predictive CDP.
We recommend you select less than 25% of the base for your volume. If the targeting volume is too high, you might target your entire active database and thus increase the chances of having too large a share of shared customers between the two volumes.
In some cases, you may need to set an audience volume lower than the volume recommended by the Splio AI, which is represented by a dip in the volume bar. If you wish to integrate this additional segment identified by our algorithms into your analysis, we recommend you read this article.
-