Katkat
Back

Design a fast host listing metrics page

AirbnbSoftware Engineer · Technical Screen
HardSoftware EngineerTechnical ScreenSystem Design

Last updated

This question evaluates backend system design and data engineering competencies, including scalable API design, time-range aggregation correctness (handling partial overlaps and state changes), data modeling and indexing, caching and precomputation, and performance validation.

Design the backend for a host (landlord) listings page.

User flow:

  • Host opens a page showing many of their listings.
  • Host selects a date range.
  • For each listing shown, the UI displays aggregated metrics for that range, e.g.:
  • Nights booked (how many nights are reserved within the range)
  • Average price within the range

Assume you start with these tables:

  • Listings(listing_id, host_id, ... )
  • Reservations(reservation_id, listing_id, start_date, end_date, status, ... )
  • Pricing(listing_id, date, price, currency, ... )

Problem:

  • When a host has 100+ listings, the page becomes slow.

Task:

  • Propose an end-to-end design (APIs, data access patterns, storage/indexing, caching/precomputation) to make the page fast and scalable.
  • Specify how you would compute the metrics correctly (including edge cases like partial overlap, cancellations) and keep them up to date.
  • Include performance goals and how you would validate them.