Coupon Experimentation Under Budget

With a fixed budget, how would you use A/B testing to find which coupon is most effective? How should consumer coupons be distributed to maximize GMV, spending the whole budget without exceeding it?

Asked in the Round 1 stage.

Address how you would compare coupon variants, define effectiveness, allocate users, calculate sample size, and control spending while preserving experimental validity.

1. State the hypothesis and choose a primary metric, guardrails, and an explicit MDE.
2. Choose the randomization unit and explain how coupon redemption and differing face values affect the design.
3. Calculate required sample size and test duration using clearly stated assumptions.
4. Describe the analysis plan, including multiple comparisons, peeking, SRM, and budget pacing.
5. Give a ship, reject, or iterate rule that maximizes GMV without exceeding the budget.