Uber
Marketplace analytics, estimation, metrics, and two-sided tradeoffs.
Working here
Working at Uber
Culture
Operational and data-heavy, with a marketplace mindset baked in: nearly every decision balances two sides (riders and drivers, eaters and couriers and restaurants). Fast and metrics-driven, with influential regional ops teams. The APM program is a structured cohort, and the analytical bar is high because the marketplace demands it.
The type of work
APMs own a piece of a two-sided marketplace and constantly reason about supply/demand tradeoffs, pricing, and incentives. The work is experiment-heavy, in close partnership with data science, ops, and eng.
A day in the life
You start in the metrics, marketplace health first: trip completion, wait times, driver utilization, and an overnight experiment readout that nudged supply the wrong way in one city. Because every lever is two-sided, the work is constant tradeoff: an incentive that pulls more drivers online this week has to be sized against what it costs and what it does to rider pricing. Mid-morning you're with a data scientist designing a switchback test, because a plain A/B leaks: treat the riders and you've changed the drivers' world too. Then an ops sync, where the regional team tells you what's actually happening on the ground and carries real weight in the decision. 'Data, not opinions' is the reflex, and you facilitate between eng, design, and sales rather than relay. You might be doing it for a market you visited on the program's research trip, the reason that city isn't just a row in a spreadsheet to you.
The interview process
Uber interview process
- 1
Recruiter screen
30 minA recruiter fit call: elevator pitch, why-Uber, your most impactful shipped feature, a level-setting behavioral. Screens motivation and communication.
Behavioral - 2
Analytical and Estimation
45 minMarket sizing, metrics, and a marketplace tradeoff.
Analytical - 3
Jam session
~45 min panelA prompt on a current Uber challenge, sent 1 to 2 days ahead. Present your framing to a PM panel, then defend it live.
Product SenseStrategyExecution - 4
Product sense
45 minDesign or improve for one side of the marketplace.
Product Sense - 5
Execution & Behavioral
45 minPrioritization under constraints and ownership stories.
ExecutionBehavioral
What they weight & how to answer
What Uber weights and how to answer
Uberrewards structured thinking over a single “right” answer. These are its strongest signals, ordered from most weighted.
Analytical
Most weightedThe signature, estimation ("ride-share trips per day in your city"), metrics, and diagnosis are weighted heavily.
This is the signature round and the bar is high because the marketplace demands it. For estimation, structure a clean top-down or bottom-up sizing and state every assumption; for metrics, reason about BOTH sides of the marketplace (riders and drivers) and the guardrail that protects the other side. Show you can move between the math and the product implication.
Product Sense
Design for a specific side of the marketplace, then name the tradeoff with the other side.
Pick one side of the marketplace to design for, then explicitly name the tradeoff you're making with the other, that two-sided reasoning is what Uber is really testing. Ground it in supply/demand or incentives, and close on a metric that captures marketplace health, not just one side's happiness.
Execution
Prioritization under real, conflicting constraints.
Expect real, conflicting constraints (ops, regional differences, two-sided incentives). Prioritize on a stated axis, sequence the work, and own the tradeoff. "Act like an owner", commit to a call rather than hedging.
Behavioral
Ownership, "act like an owner."
Ownership is the theme, "act like an owner." Bring stories where you drove something end to end and took accountability for the outcome, ideally with metrics. Keep it tight and impact-focused.
Example questions
Example Uber interview questions
“How many ride-share trips happen in your city per day?”
Estimation
“What metrics would you track for a food delivery app?”
Analytical
“Engagement on your product dropped 10% last week. How do you investigate?”
Analytical
“You have five features and time for two. How do you choose?”
Execution