Metrics & analytics · Medium
What north-star metric would you choose for a food-delivery app?
Short answer
A strong candidate is weekly orders delivered on time and as ordered. It reflects value for all three sides of the marketplace: customers get good food, restaurants get sales and delivery partners get earnings, and it leads revenue. Pair it with guardrails such as cancellation rate, delivery time and partner earnings per hour so the team cannot game it.
Detailed answer
What makes a good north-star metric
- It captures the core value users get, not just company revenue.
- It is a leading indicator of long-term business health.
- Teams can influence it through their work.
- It is hard to inflate without creating real value.
Worked example
Candidates and their weaknesses:
- Gross order value: driven up by price rises and large orders; says little about experience.
- Monthly active users: someone who opens the app and never orders still counts.
- Total orders: closer, but a late, cold or wrong order counts the same as a perfect one.
- Orders delivered on time and complete, per week: rewards both demand and fulfilment quality.
Choose the last, measured weekly because food ordering is a frequent habit and weekly cadence reduces noise from day-of-week effects.
Then break it into an input tree the teams can own:
- Active customers × orders per customer × share delivered well.
- Restaurant availability and menu accuracy.
- Delivery partner availability at peak times.
Guardrails: cancellation rate, average delivery time, customer support contacts per order, and partner earnings per active hour.
Common pitfalls
- Choosing revenue as the north star.
- Naming one metric without explaining what it ignores.
- Forgetting the supply side of a marketplace.
- Proposing a metric that takes months to move, so teams cannot learn from it.