Diagnose a 20 percent drop in activation for Transit at global scale

Problem Statement Description

Product context: Uber is a mobility and delivery platform; its products include rides, Uber Eats, grocery and retail delivery, freight, driver and courier tools, and marketplace pricing.

You are the PM responsible for Uber Transit, a rider-facing experience that helps users discover and use public transit options within the Uber app. A global dashboard shows a sudden 20% drop in Transit activation, and leadership needs to understand whether this reflects a real user behavior change, a measurement issue, a localized operational problem, or a broader product/system failure.

In this RCA interview, focus on how you would structure the investigation rather than jumping to a single cause. Activation may involve a rider meaningfully engaging with Transit for the first time, such as viewing routes, selecting a transit option, starting trip planning, or completing another defined first-use action. You should clarify the metric, isolate the anomaly, segment the drop, and identify what evidence would confirm or reject major hypotheses.

The situation is global, so your approach should account for differences across markets, rider cohorts, app versions, platforms, transit agency data integrations, seasonality, and marketplace context. You should also consider how Uber’s broader mobility ecosystem could affect Transit discovery and usage.

The experience should consider:

- How activation is defined, including numerator, denominator, eligibility, time window, and whether the metric changed recently.

- Whether the 20% decline is real or caused by instrumentation, logging, data pipeline, experiment allocation, or dashboard issues.

- Segment cuts by geography, city, platform, app version, acquisition source, rider type, language, and transit provider integration.

- Funnel analysis from Transit entry points through route discovery, option selection, and first meaningful use.

- Product, operational, and external hypotheses such as app releases, UI changes, broken transit feeds, outages, pricing/ETA changes, holidays, weather, strikes, or competitor/local mobility shifts.

- Evidence needed to prioritize hypotheses, including logs, experiment data, customer support signals, provider health metrics, and cohort trends.

- Immediate mitigation options if riders are blocked, while preserving safety, reliability, and user trust.

- Longer-term prevention through monitoring, alerting, ownership, and clearer activation health diagnostics.

Your goal is to present a clear, executive-ready RCA plan that narrows the problem from a global headline metric to testable root causes, identifies the highest-risk areas first, and explains how you would move from investigation to mitigation and prevention without assuming the answer upfront.