Support for two different versions of a design. Useberry must randomly assign the test version to the participants.In results it must be possible to compare variants and dive into the data of one variant as well.