Investigate LA Completed Orders Decline

You are a data scientist supporting DoorDash's consumer pricing team. Leadership notices that **completed orders in Los Angeles** have declined materially over the last 2 weeks relative to the prior baseline, while overall company trends are roughly stable. How would you investigate this drop? Please structure your answer to cover: 1. **Problem framing**: what clarifying questions would you ask, and how would you define the scope of the issue? 2. **Metric tree**: what primary metric, supporting metrics, and guardrail metrics would you use? Consider demand, conversion, pricing, merchant availability, delivery supply, ETAs, cancellations, and defects. 3. **Hypotheses**: list plausible explanations, including pricing changes, promotion changes, seasonality, competitor actions, Dasher supply constraints, merchant outages, product bugs, instrumentation issues, and geographic or customer-mix shifts. 4. **Analysis plan**: how would you diagnose the root cause using segmentation and causal reasoning? Discuss confounding, selection bias, and cases where aggregate trends could be misleading. 5. **Recommendations**: what short-term fixes and longer-term solutions would you propose if the issue appears to be pricing-related versus supply-related? 6. **Experimentation**: design an experiment or quasi-experiment to test a pricing intervention intended to recover completed orders. Specify the randomization unit, success metrics, guardrails, and how you would think about power, MDE, and pre-period adjustment.