When are Uber performance reviews? (2026)
Uber runs real-time marketplaces in hundreds of cities with local regulation and local economics, so results are city-shaped rather than global. Confirm cycle dates with your manager.
Twice a year
No specific review month is published for this company, because none could be sourced from the company's own material. What follows is the cycle shape, which is what is actually knowable. Your own manager is the authority on this year's dates.
How the cycle works
Uber operates a real-time matching problem across hundreds of cities, each with its own regulation, its own supply economics, and its own demand pattern. That geographic fragmentation is the single most important thing to understand about how results are judged here. A global average is nearly meaningless: a pricing change can work in one city and fail in the next one over, and a company-wide number can conceal both.
The real-time constraint is the second structural fact. Decisions are made in seconds with incomplete information, at very high volume, and the systems doing it cannot pause to think. That puts an unusual premium on latency, on graceful degradation, and on systems that behave sanely when a dependency is slow.
Experimentation is heavily used, which means most claims can and should be backed by an experiment. An unquantified claim in a company this instrumented reads as work that was never measured.
Cycle timing varies by organization. Ask your manager.
What actually gets weighed
Marketplace health, reported by city or by cohort rather than globally: completion rates, wait times, earner utilization, and the balance between them. The disaggregation is the honesty.
Real-time system performance is the second axis. Latency at the tail, behavior under partial failure, and the ability to keep matching when something upstream degrades.
Earner outcomes are weighted as a distinct constituency. Utilization and earnings stability for drivers and couriers are their own results, not a footnote to rider experience.
Regulatory and city-launch work is the fourth. Operating legally in a new jurisdiction is a genuine cross-functional accomplishment involving product, legal and operations together, and it is easy to undersell as coordination.
What to have ready
- Results broken out by city or cohort, since that is the only level at which they mean something.
- Experiment design and outcomes, including the ones that did not work.
- Latency and degradation behavior for anything on the real-time path.
- Earner-side outcomes stated separately from rider-side outcomes.
- Regulatory or launch work, with the jurisdictions named and your specific part in it.