Build a metric tree for Experiences after a major redesign

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, changing how travelers discover, evaluate, and book activities hosted by locals, as well as how hosts present and manage their offerings. You are asked to build a metric tree that helps the team understand whether the redesign is improving the health of the Experiences marketplace.

This is a metrics interview focused on defining a structured measurement system, not proposing the redesign itself. Your metric tree should connect high-level business and user outcomes to measurable drivers across the traveler journey, host supply health, marketplace quality, trust, and operational reliability.

The experience should consider:

- A clear top-level success metric and how it reflects value for Airbnb, guests, and Experiences hosts

- Denominators and definitions for key rates, such as discovery, booking, completion, cancellation, and repeat behavior

- How to separate guest-side demand metrics from host-side supply and quality metrics

- Instrumentation needed across search, listing views, booking flow, payments, host tools, reviews, and support events

- Cohorts such as new vs. returning guests, new vs. established hosts, geography, trip type, and experience category

- Guardrail metrics for safety, trust, cancellations, refunds, disputes, host churn, and marketplace quality

- How to distinguish redesign impact from seasonality, destination mix, marketing, pricing, or supply changes

- How the metric tree would be used to diagnose whether a change is caused by awareness, conversion, availability, quality, or retention

The goal is to create a decision-useful metric framework that Airbnb’s product, design, data, and operations teams could use after launch to monitor marketplace health, identify where the redesign is helping or hurting, and decide what areas need deeper investigation.