Investigate why conversion fell after a Host Tools launch 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 new set of Host Tools globally, intended to help hosts manage listings, availability, pricing, amenities, and guest expectations more effectively. Shortly after launch, the marketplace team observes a meaningful decline in conversion, particularly affecting the journey from search or listing view to completed booking. You are asked to investigate what may have happened.

This is a root-cause analysis problem in a global two-sided marketplace. The launch may have changed host behavior, listing quality, price presentation, availability, ranking signals, guest trust, or the booking flow itself. The issue may also vary by geography, device, property type, trip length, host segment, or traveler segment such as remote workers looking for longer stays, strong Wi-Fi, workspace amenities, and flexible cancellation.

Assume the drop is significant enough to require urgent investigation, but the exact cause is unknown. Your task is to structure how you would diagnose the anomaly, separate product-caused impact from external factors, identify the most likely drivers, and recommend what evidence would be needed before taking mitigation action.

The experience should consider:

- How “conversion” should be defined across the Airbnb funnel, including search-to-listing, listing-to-booking-start, booking-start-to-confirmed-booking, and host acceptance where relevant.

- Whether the decline is global or concentrated by market, platform, guest segment, host segment, listing type, trip duration, price band, or acquisition channel.

- Instrumentation checks to confirm the metric drop is real and not caused by tracking, logging, attribution, experiment exposure, or data pipeline issues.

- How the Host Tools launch could have affected host-side inputs such as availability, minimum nights, pricing, fees, amenities, cancellation policies, response time, or listing completeness.

- How guest-facing surfaces may have changed indirectly, including search ranking, listing detail pages, price transparency, trust cues, amenity filters, or booking friction.

- External and marketplace factors to rule out, such as seasonality, competitor actions, travel demand shifts, local regulations, supply constraints, or macro events.

- Evidence needed to prioritize hypotheses, including cohort cuts, pre/post comparisons, experiment holdouts, funnel diagnostics, host behavior logs, guest feedback, and support/contact patterns.

- Mitigation and prevention considerations, including whether to pause, roll back, target fixes by segment, communicate with hosts, or add monitoring for future host-tool launches.

The goal is to demonstrate a rigorous RCA approach: frame the anomaly clearly, validate the data, segment the impact, generate and test plausible hypotheses, identify the most likely root cause, and outline responsible next steps that protect guest trust, host supply health, and marketplace conversion at global scale.