Zomato: Make Restaurant Scores Trustworthy and Useful Across Dine-in and Delivery
Company: Zomato
Role: Product Manager
Category: Product Design & Strategy
Difficulty: medium
Interview Round: Other
You are a product manager at Zomato, where diners rate restaurants both for eating in and for delivery orders. How would you improve restaurant ratings so they work well for both? Identify the problems worth solving, decide whom the changes are for, describe what you would change, and explain how you would know it worked.
```hint What one number hides
A delivery experience depends on the kitchen, the packaging and the delivery leg, while a dine-in visit also depends on service and ambience. Start from the decision a rating is meant to support, then ask what a single score per restaurant hides or gets wrong.
```
### Clarifying Questions
- Does the app currently show one combined score per restaurant, or separate dining and delivery scores?
- Is the main goal better choices for diners, more trust in ratings, or more useful feedback for restaurants?
- What prompted this: low rating volume, suspected fake ratings, complaints from restaurants, or ratings that do not match what diners experience?
- Is the scope the score only, or written reviews and photos as well?
### What a Strong Answer Covers
- A clear purpose for ratings and the users they serve: diners choosing, restaurants improving, and the app's own ranking
- Specific weaknesses of restaurant ratings across dining and delivery, such as delivery problems landing on the restaurant, sparse or stale dining ratings, selection bias and manipulation
- Prioritized changes to how ratings are collected, computed and displayed, with effort and risk weighed
- Success metrics that test whether ratings predict the diner's experience, with guardrails for rating volume and fairness to restaurants
- Risks such as gaming, restaurant pushback and extra friction for people leaving ratings
### Follow-up Questions
- A restaurant says its delivery rating is low because of late deliveries it does not control. How does your design handle that?
- How would you keep ratings from inflating until most restaurants sit in a narrow band near the top?
- How should the app rate a restaurant that gets many delivery orders but very few dine-in visits?
Overview: A product design prompt about improving Zomato's restaurant ratings so they serve both dine-in visits and delivery orders. It tests diagnosing rating problems such as delivery issues blamed on restaurants, bias and stale scores, redesigning how ratings are collected and computed, and measuring whether ratings predict diner experience.