Whenever you're ready, walk me through how you'd approach this.
What methods would you use to assess the statistical significance of your A/B test results?
Explain how you would choose an appropriate statistical test, define the null and alternative hypotheses, and report uncertainty around the estimated treatment effect. Include how you would account for sample size, minimum detectable effect, multiple comparisons, peeking, and differences between the randomization and analysis units. Describe how you would distinguish statistical significance from practical significance and how you would decide whether to ship.