Shared-Ride Trade-offs at Uber: Riders, Drivers and Matching Rules for UberPool

You are a PM at Uber designing UberPool, the option where riders heading in a similar direction share one car and each pays less than for a private ride. What parameters would you consider in its design? ```hint Name what each rider gives up A shared trip is a bargain: a lower fare in exchange for something else. Work out what each rider and the driver trades away, and which settings in the matching system decide how much they trade. ``` ### Clarifying Questions - By "parameters", do you mean the rider, driver and business factors the design must balance, the tunable settings of the matching system (detour limit, pickup wait, seats per car, discount), or both? - Is this a new launch in cities that only offer private rides, or a redesign of an existing shared option? - Which goal leads: cheaper trips for price-sensitive riders, more paid trips per driver hour, or fewer cars on the road at peak? - Which trips are in scope: weekday commutes, late-night rides after events, airport runs, or all of them? ### What a Strong Answer Covers - The goal of shared rides for riders, drivers and Uber, and the rider and driver segments chosen to design for first - The rider and driver factors at stake, including safety, ranked for the chosen segment - The matching settings that encode those factors, how they pull against each other, and the role of rider density - Success metrics with guardrails for reliability, safety, driver earnings and cannibalization of private rides - Edge cases such as no match found, co-rider no-shows, group bookings and luggage ### Follow-up Questions - Outside peak hours in a smaller city, most shared requests find no match. What would you change? - Riders complain that shared trips often arrive much later than the estimate shown at booking. Which parameter failed, and how would you fix it? - How would you set the size of the shared-ride discount relative to a private ride?