Explain the technical trade-offs of adding AI capabilities to Ads at global scale

Problem Statement Description

Product context: DoorDash is a local commerce and delivery platform; its products include restaurant delivery, DashPass, grocery and retail delivery, merchant tools, and dasher tools.

DoorDash is exploring how to add AI capabilities to its Ads products at global scale across local commerce categories such as restaurants, grocery, convenience, and retail. These capabilities could affect how merchants create campaigns, how ads are targeted or ranked, how budgets are optimized, how creative is generated, and how consumers discover sponsored options in suburban and other markets.

In this technical PM interview, you are asked to explain the technical trade-offs involved in bringing AI into a large-scale advertising system. The context includes a multi-sided marketplace where consumers expect relevant choices, merchants expect measurable ROI, Dashers are affected by demand patterns, and DoorDash must protect delivery reliability, unit economics, and marketplace balance.

Your discussion should focus on the product and technical implications rather than proposing a full implementation plan. Consider what changes when AI systems operate across many geographies, merchant types, catalog structures, consumer behaviors, regulatory environments, and latency-sensitive ordering flows.

The experience should consider:

- Core AI use cases for Ads, such as targeting, ranking, bidding, budget pacing, creative generation, recommendations, or merchant-facing campaign automation

- Data requirements across consumers, merchants, menus/catalogs, orders, locations, delivery constraints, ad interactions, and conversion events

- API and system integration points with ad serving, search, feed ranking, checkout, merchant tools, experimentation, billing, and reporting

- Latency, reliability, fallback behavior, and uptime expectations for ads embedded in real-time consumer experiences

- Privacy, security, consent, data retention, and regional compliance considerations for using consumer and merchant data

- Model quality trade-offs, including relevance, fairness, explainability, cold-start handling, hallucination risk, and bias across suburban versus dense urban markets

- Rollout, experimentation, observability, monitoring, incident response, and human review needs for AI-powered ad decisions

- Business trade-offs across merchant ROI, consumer trust, ad revenue, delivery operations, Dasher earnings, and long-term marketplace health

The goal is to demonstrate how you reason as a Technical PM about adding AI to a global ads platform: clarifying requirements, identifying architectural and data dependencies, evaluating risks and trade-offs, and balancing product impact with operational robustness, compliance, and marketplace trust.