Investigate why conversion fell after a Merchant Portal launch 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 has launched a new version of its Merchant Portal globally for restaurants, grocery, convenience, and other local commerce partners. Shortly after launch, the team observes a meaningful drop in conversion within the portal. This may affect merchants completing critical workflows such as onboarding, menu updates, store activation, promotion setup, ads enrollment, payout setup, or other revenue-impacting actions.

You are the PM asked to lead the root-cause investigation. The situation is high stakes because the Merchant Portal is a core operating surface for merchants, and friction can hurt merchant ROI, marketplace supply, consumer selection, Dasher earnings, and DoorDash unit economics. The issue is global, so the investigation must account for differences across countries, languages, merchant segments, verticals, devices, browsers, permissions, and rollout cohorts.

Frame how you would diagnose the anomaly, validate whether the conversion decline is real, identify likely causes, decide what to do immediately, and prevent similar issues in future launches. Do not jump directly to a fix; focus on the investigation structure, evidence needed, and decision-making process.

The experience should consider:

- The exact conversion metric, funnel steps, denominator, time window, and expected baseline before and after launch

- Whether the drop is isolated to specific merchant segments, geographies, verticals, account types, devices, browsers, or user roles

- Instrumentation checks, event logging changes, attribution gaps, experiment assignment, and data pipeline reliability

- Product workflow changes that may have introduced friction, confusion, latency, permission issues, localization problems, or broken states

- Operational and launch factors such as phased rollout, account-manager communications, training, support tickets, and release timing

- Hypotheses spanning UX, technical reliability, policy/configuration changes, pricing or incentive changes, and external marketplace conditions

- Evidence needed to prioritize hypotheses, including funnel analytics, session replays, merchant feedback, support contacts, error logs, and cohort comparisons

- Immediate mitigation options, rollback or holdout considerations, monitoring needs, and longer-term prevention mechanisms

Your goal is to present a rigorous RCA approach that helps DoorDash quickly determine whether the conversion decline is caused by measurement, product experience, technical defects, rollout execution, or merchant behavior changes, while protecting merchant trust and marketplace health.