Give an example of a site change that we can't test on a subset of users via a controlled experiment. How would you estimate the impact of that change?
Lesson content locked
If you're already enrolled, sign in. Otherwise, enroll to unlock the lesson.