One way to measure whether a team’s cumulative effect is positive is to have a long term holdout of users who never see their experiments. You can check for a s

One way to measure whether a team’s cumulative effect is positive is to have a long term holdout of users who never see their experiments. You can check for a statically significant difference between this holdout group and the group of users who see the features the team shipped.