The structured answer:

Candidate metrics and their failure modes:

Total rides completed — easy to measure, but rewards volume without regard to profitability. Surge-induced cancellations count as 0. Doesn't capture driver side.
Gross bookings ($) — incorporates price, but still single-sided. Also exposed to macro shifts (gas prices, holidays) that aren't the team's work.
Monthly active riders × rides per rider — better, because it separates the acquisition story from the engagement story. But doesn't include driver supply.
Completed rides per active rider — engagement metric. Good for understanding habit, bad as a North Star because a small, hyper-active user base would win.
Driver utilization (% of driver online time spent on rides) — critical supply-side metric. But doesn't directly measure rider experience.
Two-sided matches — riders who got a ride in X minutes × drivers earning above Y — captures both sides of the marketplace. Complex to compute and communicate.

A defensible North Star proposal:

"I'd propose completed rides per active rider per month, with guardrails on driver earnings per hour and rider wait time. The metric rewards engagement (rides per rider), user retention (rides is 0 for churned users), and marketplace health (can't grow it without supply). The guardrails make sure we can't buy the metric by screwing drivers (driver earnings floor) or by forcing wait times down (riders tolerate 8 min, not 15)."

Why guardrails matter: the failure mode of any single metric is that teams optimize it in ways that break other things. At Uber, "rides completed" without wait-time guardrails could be bought by matching riders with drivers 30 minutes away. Having guardrails makes the metric gameable in principle but not in practice.

What senior answers add:

Acknowledge tradeoffs — a single North Star can't capture everything; you're picking the one thing executives will actually look at daily and letting guardrails catch pathology.
Reference historical wins — Airbnb famously chose nights booked over signups; Meta chose daily active users over total users. Both bets reflected a belief about what would predict long-term value.
Propose a sunset plan — the North Star at a $1B company isn't the right one at $100B. Build in a review cadence.

What's the North Star metric for a ride-share company?

Draft your answer

Stuck? Peek at a hint

More product metrics questions