Evaluate Growth and Pricing for a Grocery Delivery Startup
Capital OneData Scientist · Technical Screen
Last updated
Evaluate a grocery delivery platform's unit economics across two metro markets, construct an end-to-end marketplace metric framework, and determine whether a proposed dynamic delivery fee increases contribution profit.
Scenario
You are embedded as a data scientist in a fast-growing on-demand grocery delivery startup. Leadership is considering a new pricing strategy: shifting from a flat delivery fee of $3.99 per order to a distance-and-demand tiered fee averaging $5.49 per order. You have pilot test data from two markets over the last 30 days: Market A (dense urban) and Market B (suburban spread).
Pilot Data
During a 30-day A/B test, users in each market were randomly assigned to either the Control (flat $3.99 fee) or Treatment (tiered pricing averaging $5.49 fee).
| Market | Group | Orders | Avg Gross Merchandise Value (GMV) | Platform Commission Rate | Courier Payout per Order | Payment Processing Fee | Customer Support / Refund Cost |
|---|---|---|---|---|---|---|---|
| Market A | Control | 40,000 | $62.00 | 12% | $7.20 | 2.5% of GMV + $0.30 | $1.10 per order |
| Market A | Treatment | 38,000 | $66.00 | 12% | $7.20 | 2.5% of GMV + $0.30 | $1.15 per order |
| Market B | Control | 25,000 | $84.00 | 10% | $9.80 | 2.5% of GMV + $0.30 | $1.60 per order |
| Market B | Treatment | 21,500 | $91.00 | 10% | $9.80 | 2.5% of GMV + $0.30 | $1.85 per order |
Note: Platform Net Revenue per order is defined as (Commission Rate * GMV) + Customer Delivery Fee.
Contribution Profit per order is defined as Net Revenue - Direct Variable Costs (Courier Payout + Payment Processing + Support/Refunds).
Your Task
- Metric Hierarchy: Design a three-tier metric framework (North Star, Primary / Operational metrics, Guardrail metrics) to monitor marketplace health, distinguishing GMV from Platform Net Revenue.
- Unit Economics & Total Contribution Profit:
- Compute Platform Net Revenue per order and Contribution Profit per order for Control vs. Treatment in both markets.
- Calculate the total 30-day Contribution Profit change in dollars and percentage for each market.
- Market Comparison & Elasticity: Explain why the change in order volume and GMV differed between Market A and Market B. What does this indicate about customer price elasticity and order basket behavior across urban vs. suburban settings?
- Recommendation & Stress Testing: Would you roll out this pricing change to either or both markets? What key operational risks or courier-side supply dynamics could overturn your recommendation over a 6-month horizon?
Hints
Hint 1
Start by writing out the exact formula for Net Revenue and Variable Cost per order before aggregating to the total market level.
Hint 2
Pay attention to the interaction between order loss and basket size (GMV) increase: higher basket size increases both commission revenue and variable processing fees.
Hint 3
Consider courier retention and vehicle utilization: if order volume drops significantly in suburban markets, couriers experience longer idle times between dispatches.
Follow-up questions
- How would you isolate whether the GMV increase in Treatment was caused by users consolidating baskets to justify the higher delivery fee versus churn of lower-income, small-basket users?
- If the merchant commission structure changed from a flat percentage to a tiered take-rate based on merchant volume, how would your unit economics calculation adapt?
- How would you design a synthetic control or difference-in-differences test if city-wide network effects prevent user-level A/B randomization?
Solution walkthrough is for members
Get the full walkthrough with reference code and complexity analysis, plus Kitty Coach at 300 messages a day. Hints, the code runner and community answers stay free.