Describing how each test variant is different would improve the experience of comparing performance. Ideally a name could be set during variant creation, which is then shown in the UI and written to the event stream available through tools like Segment