Build a metric tree for Experiences after a major redesign at global scale

Problem Statement Description

Product context: Airbnb is a travel marketplace connecting guests and hosts; its products include stays, experiences, host tools, guest booking, trust and safety, payments, and reviews.

Airbnb has recently launched a major redesign of its Experiences product globally. The redesign may affect how guests discover and book activities, how hosts create and manage listings, and how the marketplace balances quality, trust, availability, and demand across different regions.

You are being asked to define a metric tree that helps Airbnb understand whether the redesigned Experiences product is succeeding. The focus should include both guest-side outcomes and host-side marketplace health, with particular attention to global scale, regional differences, and the operational realities of running an experiences marketplace.

This is a metrics interview question, so the emphasis is not on proposing new product features. Instead, clarify what success means, how it should be measured, how metrics connect to one another, and how Airbnb should use the metric tree to make decisions after the redesign.

The experience should consider:

- A clear north-star or top-level success metric for Airbnb Experiences, including what is counted and what is excluded.

- Guest funnel metrics across discovery, listing evaluation, booking, attendance, satisfaction, and repeat usage.

- Host-side metrics such as active supply, listing quality, availability, earnings, cancellations, response behavior, and retention.

- Marketplace balance metrics that capture supply-demand fit, conversion, utilization, pricing, geographic coverage, and seasonality.

- Metric definitions with appropriate denominators, time windows, and handling of edge cases such as cancellations, refunds, reschedules, no-shows, and multi-person bookings.

- Cohorts and cuts by geography, traveler type, host maturity, experience category, acquisition channel, device, and pre- vs. post-redesign exposure.

- Instrumentation needs for the redesigned flows, including event tracking, attribution, funnel continuity, and data quality checks.

- Guardrail metrics around trust, safety, customer support contacts, disputes, review quality, host churn, guest dissatisfaction, and marketplace degradation.

The goal is to produce a structured metric tree that Airbnb leadership, product teams, data science, and operations teams could use to evaluate the redesign, diagnose performance changes, and decide where to investigate or intervene without relying on a single vanity metric.