You're planning an experiment on a product change and need to decide whether the test is worth running. Before launch, you want to know how much traffic you need and what size effect the experiment can realistically detect.
How would you think about sample size and power before launching an experiment?