Define the North Star metric and guardrails for restaurant discovery and review trust

Problem Statement Description

Zomato helps urban diners decide where to eat or order from by combining restaurant listings, menus, ratings, reviews, photos, delivery options, table booking, offers, and sponsored placements. For diners with dietary constraints—such as vegetarian, vegan, Jain, halal, gluten-free, allergies, or health-driven preferences—the discovery journey depends heavily on trust: whether restaurant information is accurate, reviews are credible, menus are current, and filters actually surface safe and relevant choices.

In this metrics interview, you are asked to define a North Star metric for restaurant discovery and review trust, along with guardrail metrics that prevent the product from optimizing only for short-term conversion. The metric system should connect user value, trusted decision-making, and business outcomes such as booking and order conversion, while accounting for review quality, sponsored ranking pressure, restaurant data freshness, privacy, accessibility, and operational cost.

Your scope is the discovery-to-decision journey: searching or browsing restaurants, applying dietary filters, reading reviews, comparing options, and then placing an order, booking a table, saving, calling, or otherwise taking a high-intent action. You should clarify what “success” means, who is included in the denominator, how the metric would be instrumented, and how the metric would remain useful across cohorts, cities, cuisines, restaurant types, and user intent.

The experience should consider:

- A clear North Star metric definition, including numerator, denominator, time window, and why it reflects trusted restaurant discovery.

- How to distinguish meaningful successful outcomes from low-quality clicks, accidental taps, ad-driven visits, or shallow engagement.

- Cohorts such as new vs returning users, dietary-constrained diners vs general diners, high-frequency vs occasional users, and city or cuisine-level differences.

- Instrumentation needed across search, filters, restaurant detail pages, review consumption, menu views, booking/order actions, and post-visit or post-order feedback.

- Guardrails for review trust, misinformation, fake or incentivized reviews, rating inflation, stale menus, incorrect dietary tags, and ranking fairness.

- Business and ecosystem guardrails, including restaurant partner satisfaction, ad load, conversion quality, cancellations, refunds, complaints, and operational burden.

- User protection guardrails around privacy, sensitive preference handling, accessibility, and responsible personalization.

- Decision usefulness: how the metric would guide ranking, review quality investments, restaurant data operations, product experiments, and trade-offs between revenue and trust.

The goal is to define a practical metrics framework that helps Zomato measure whether diners are confidently finding restaurants they can trust, especially when dietary needs raise the cost of a bad decision, while ensuring that growth in bookings or orders does not come at the expense of review integrity, user safety, or long-term marketplace trust.