How would you improve reliability and latency for Ads

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.

Uber operates a global marketplace where ads can appear across rider, eater, courier, and driver-facing surfaces. In this question, focus on Ads experiences relevant to drivers, such as sponsored earning opportunities, merchant or brand promotions, in-app placements, or marketplace messages that must load quickly and reliably without disrupting core trip, delivery, navigation, or earnings workflows.

You are being asked to think like a Technical PM responsible for improving the reliability and latency of the Ads platform. The challenge is not just to make ads faster in isolation, but to define what “reliable” and “low latency” mean across different driver contexts, geographies, network conditions, app states, and marketplace moments where missed, delayed, or broken ads can affect revenue, trust, and driver experience.

Your response should frame the product and technical problem, identify the systems and stakeholders involved, and explain how you would prioritize improvements while balancing monetization, driver safety, marketplace health, privacy, and engineering cost.

The experience should consider:

- Driver workflows where ads may appear, including online/offline states, trip acceptance, waiting time, delivery pickup/dropoff, earnings, and incentives-related screens.

- Reliability and latency definitions, such as ad request success rate, render success, time to first ad, timeout behavior, freshness, error rates, and regional/network variance.

- APIs, data flows, and dependencies across ad serving, targeting, ranking, budgeting, experimentation, client rendering, caching, and observability systems.

- Failure modes, including slow network, backend timeout, invalid creative, targeting service degradation, stale campaigns, client version fragmentation, and third-party dependency issues.

- Privacy, safety, and compliance constraints around driver data usage, location, personalization, frequency capping, and sensitive in-trip contexts.

- Trade-offs between ad revenue, driver experience, app performance, marketplace reliability, and operational complexity.

- Rollout and experimentation considerations, including cohorts, markets, app versions, guardrails, fallbacks, and rollback criteria.

- Monitoring and incident response needs, including dashboards, alerts, SLOs, root-cause workflows, and post-launch validation.

The goal is to describe how you would approach improving Ads reliability and latency in a technically rigorous, product-aware way, showing how you would define success, identify bottlenecks, align teams, and make trade-offs appropriate for Uber’s real-time marketplace environment.