Prioritize a new warehouse proposal with data
Company: Amazon
Role: Data Scientist
Category: Analytics & Experimentation
Difficulty: hard
Interview Round: Technical Screen
Stakeholders ask whether to build a new fulfillment center (FC). You have partial data: daily demand by metro, current FC capacities and per-FC variable costs, average delivery time vs. distance, and lease/CapEx options. 1) Outline the end-to-end analysis plan to recommend build vs. lease vs. defer: demand forecasting method, service-level targets, network optimization objective (e.g., minimize total cost subject to 95% next-day coverage), and key constraints. 2) Show a back-of-the-envelope model with explicit assumptions to estimate NPV over 5 years, including cannibalization and ramp curves. 3) Describe the experiment or quasi-experiment you would run before committing (geo holdouts, incrementality via synthetic control, or simulation with historical replays), with success metrics and stopping rules. 4) Provide a sensitivity analysis: which three parameters most threaten the decision and how you would bound them with additional data. 5) Draft a one-slide stakeholder narrative with the recommendation and the top two risks plus mitigations.
Quick Answer: This question evaluates demand forecasting, network optimization, five-year NPV financial modeling, experiment and quasi-experiment design for incrementality estimation, sensitivity analysis, and stakeholder communication within the analytics & experimentation domain for data science roles focused on warehouse/fulfillment center decisions.