# ask-questions


02/29/2024, 12:10 PM
Hi! I’m setting up a scenario like this: • We are releasing a new feature for which we want to start with a rollout to 10% of the users • Inside the feature we want to run an AB test for 2 modals To do this we have created 2 feature flags, 1 for the feature itself and one for the AB test which is implemented dependant on the activation status of the first one. In terms of metrics measurement: • For the the AB test we should be ok, since the rule it’s connected to the experiment created • We have a doubt on the rollout rule: As per you docs -> Percentage Rollouts do not fire any tracking calls so there’s no way to precisely correlate the rollout to changes in your application’s metrics Which means that we cannot track the result of the new feature against the current one Should it be better to: • Transform the rollout rule to AB test 90/10 • Increase the rollout percentage to 20% and put an AB test 50/50, so to have 10% of the users actually seeing the new feature, and tracking the other 10% with the current one? Any suggestion?