Has there ever been discussions about separating experiments with feature rollouts with metrics? The scenario is when we want to rollout a feature and see how it impacts certain metrics I don't consider it really as an experiment. Doing it the experiment way you also need to change the traffic included in the experiment to whatever percentage you want first and then the split to 100% for the variation. Would be nice to just be able to add metrics to the percentage rollout rule since you already have to enroll based on attribute. Doesn't clutter the experiment page either with non experiments.

Something like this is in the works in the near term!

Absolutely fantastic! Thank you 😄