Does Evidently keep users in the same variant group throughout the experiment

Question

We are looking into using Evidently for A/B testing but during trials have found that I could be assigned to one group but a little bit later would be assigned to another.
I don't see info in the documentation stating explicitly that users will be assigned to a certain group throughout the whole duration. Does anybody here know the technicality of it?

Answer

If a feature is launched, users could be placed in a variant. This depends on the allocation percentage. If they're part of the launch, their variant depends on the feature's variations. 
If not part of a launch, they might be in an experiment variant, depending on its allocation. 
If not in either, they get the default variant.

[+] https://docs.aws.amazon.com/cloudwatchevidently/latest/APIReference/API_BatchEvaluateFeature.html

Does Evidently keep users in the same variant group throughout the experiment

Contenus pertinents