The A/B test in Datatrics is really meant to measure the effect of personalization on your website. As time goes on and it is clear that personalization works for you, it is also advisable to make the control group smaller and smaller. This will allow you to benefit more from the influence of personalization on your target audience, as it will result in higher sales.

A/B testing, for example, two different buttons with an external tool (like Google Optimize) is something that can simply be done simultaneously. However, it is advisable to put the different groups side by side and see if the proportions are right.

Suppose the A/B test in Datatrics is still at 50/50 and you also start an A/B test to test two variants of your shopping cart page, then of course the desired situation is:

25% see variant A of the shopping cart and see Datatrics personalizations

25% see variant B of the shopping cart and see Datatrics personalizations

25% see alternative A of the shopping cart and don't see any Datatrics personalizations

25% sees variation B of the shopping cart and sees no Datatrics personalizations

In practice, however, this can also turn out differently and you will see other proportions.

As you may know, in Google Analytics you can segment the two Datatrics A/B groups, for many genuine A/B testing tools this is also possible. We then always recommend creating the four groups named above as segments in Google Analytics and keep an eye on whether the proportions (remain) correct. You can then compare group 1 with group 2, but even more important to find out the effect of just the shopping cart A/B test: compare groups 3 & 4.

By comparing these groups you know for sure that you only measure the effect of the shopping cart A/B test or other A/B tests in the future.