Metrics question

You are the PM for ad products within Meta, specializing in offline shops. How will you measure the success of the product?

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

Metrics design for a hard to measure ad product: can you connect an online ad to real world, offline purchase outcomes.

State the core challenge: offline shop ads are hard to measure since the purchase happens outside the app, unlike a click to purchase online funnel.
Propose proxy metrics: store visit rate, measured through location data or check ins, and offline conversion lift through matched control groups.
Add a second proxy, coupon or promo code redemption in store tied to the ad exposure, since it gives a directly attributable signal.
Set advertiser side success as return on ad spend calculated from estimated offline lift, not just impressions or clicks.
Add a guardrail: user trust, tracked through opt out rate of location sharing, since offline attribution requires more invasive data.
State you would confirm with the interviewer what data access, like location permissions, is realistically available.

Addresses the core measurement problem directly, offline attribution, rather than defaulting to online metrics like clicks that do not fit the product.
Proposes a specific attribution method, matched control groups for visit lift, showing real analytical depth.
Flags privacy and trust as a guardrail, since offline tracking is more sensitive than online clicks.

Measuring only impressions and clicks, which do not capture the actual offline outcome advertisers care about.
Ignoring the privacy sensitivity of location based attribution.

How would you prove the ad caused the visit rather than just correlating with it?
How would you handle advertisers who do not want to share redemption data?