Plan parking demand for a retail store
Company: Samsung
Role: Data Scientist
Category: Analytics & Experimentation
Difficulty: medium
Interview Round: HR Screen
A retailer similar to Best Buy is considering building a customer parking lot next to one of its stores. You are asked to determine what data you would collect and how you would use it to recommend the right parking capacity.
Please describe:
- the business objective, such as improving customer convenience and sales while controlling construction and operating cost;
- the key metrics you would estimate, for example peak parking occupancy, overflow rate, customer search time, lost visits due to insufficient parking, and return on investment;
- the internal and external data you would gather, such as store foot traffic by hour, transaction counts, average basket size, customer arrival mode split, dwell time, employee parking demand, seasonality, nearby competitor activity, local events, weather, public transit access, and land or zoning constraints;
- how you would model parking demand over time and size the lot under uncertainty;
- what confounders or biases could mislead the analysis;
- how you would validate the recommendation, for example using historical data, temporary overflow parking, or a pilot.
Discuss the tradeoff between underbuilding the lot and overbuilding it.
Quick Answer: This question evaluates demand forecasting, capacity planning, causal inference, and experimental-design competencies within data science, focusing on translating business objectives into measurable metrics such as peak occupancy, overflow rate, lost visits, and return on investment.