Vibe Engines
YouTube
System Design

Design Airbnb

Step 1 / 9

Learn system design by building a lodging marketplace like Airbnb step by step.

The numbers to beatquadtreespatial splitviewportbounded querymsgeo lookup

The whole design, in writing

Learn system design by building a lodging marketplace like Airbnb step by step. An interactive guide covering listings, geo search with filters, availability calendars, booking without double-booking, the payment hold-and-capture flow, reviews and reindexing, and scaling reads.

Every step of the build above, written out: the problem each piece solves, the option that was taken and the ones that were not, the numbers, and how it fails in production.

The big idea

What is Airbnb?

A two-sided marketplace: hosts list homes, guests search by place and dates and book a stay. The reads (search) are enormous and demand geo + availability filtering; the writes (bookings) are rare but absolutely must never double-book a home.

Guestsearch · book
New in this step: Guest.

Build a fast, filterable search over a spatial index and availability calendars, then a transactional booking path that reserves dates atomically and holds payment. Search optimizes for scale; booking optimizes for correctness.

What the new pieces do

Guestclient
Searches by place and dates, browses listings, and books a stay. Search is the heavy, read-mostly path; booking is rare but must be rock-solid.

Step 1 · The skeleton

List and find homes

Hosts need to publish listings and guests need to find them. Even ignoring geography and dates, you need a clean split between the catalog of homes and the act of searching it.

GuestHostAPI GatewaySearch ServiceListings DB
New in this step: Host, API Gateway, Search Service, Listings DB. · swipe to pan the diagram

Guest searches massively outnumber host writes. How do you structure search vs the catalog?

  1. Heavy search load hammering the source-of-truth DB threatens the writes (new listings, bookings) and can’t scale independently. Search should not contend with the system of record.

  2. The gateway routes host writes to the Listings DB and guest queries to a search service reading an index derived from it. Search scales for read traffic while writes stay simple — and a search overload never threatens the source of truth.

  3. You can’t pre-cache the combinatorial explosion of place × dates × filters, and listings/availability change constantly. You need a live searchable index, with caching layered on later for hot queries.

An API Gateway sends host writes to a Listings DB and guest queries to a Search Service that reads from it. This separation lets search scale for heavy read traffic while writes stay simple and consistent.

What the new pieces do

Hostclient
Publishes a listing with photos, price and an availability calendar. Writes are infrequent compared to the flood of guest searches.
API Gatewaybackend
The single entry point routing searches to the search service and bookings to the booking service, behind a load balancer.
Search Serviceservice
Finds listings matching a location, date range, price and amenities, ranked for relevance. The most-hit service in the whole system.
Listings DBstore
The source of truth for listings: descriptions, photos (in blob storage), pricing rules and host info. Read constantly during search.

Step 2 · Search by place

Geo indexing

“Homes near this map area” is the core query, but computing distance to every listing on earth per search is impossibly slow. A normal database index on lat/long doesn’t handle 2D proximity well.

Search Servicegeo + filtersGeo Indexquadtree / geohash
New in this step: Geo Index.

"Homes near this map area" is the core query. How do you avoid computing distance to every listing on earth?

  1. A 1-D index can’t answer 2-D proximity — nearby in one axis can be far in the other, so you’d still scan and distance-check huge ranges. 2-D search needs a spatial structure.

  2. Distance from one origin doesn’t help search from arbitrary map locations and must be recomputed per query. You need a structure indexed by region, not by distance to one point.

  3. Geohashing turns 2-D coordinates into a sortable prefix; quadtrees split dense areas. Proximity search becomes a bounded lookup of listings in the viewport, then filter/rank that small set.

Build a spatial index — a quadtree or geohash — that buckets listings by region so you can fetch just the ones inside the viewport quickly, then filter and rank that small set.

  • quadtreespatial split
  • viewportbounded query
  • msgeo lookup

What the new pieces do

Geo Indexindex
A spatial index that answers “listings within this map area” quickly, instead of scanning every home on earth and computing distances.

Back of the envelope

geohash = sortable prefix
2-D coords → a 1-D bounded range
quadtree splits dense areas
a viewport query, not a planet scan
fetch in-viewport, then filter/rank
work on a small candidate set

Step 3 · Only what’s free

Availability calendars

Guests search by date range, so results must exclude homes already booked for those nights. Availability changes constantly as bookings happen, and it gates both search and booking.

Search Servicegeo + filtersGeo Indexquadtree / geohashListings DBhomes · pricesAvailabilitycalendar / listing
New in this step: Availability.

Guests search by date range, so results must exclude homes booked for those nights. How do you model "is it free?"

  1. A home is free some nights and taken others — a single flag can’t express a calendar. A stay spans a contiguous range of nights that must all be free.

  2. One authoritative availability calendar per listing lets search filter by the requested range and booking later reserve against the same calendar. Modeling a date range makes both filtering and the no-double-book check natural.

  3. Scanning and aggregating all bookings per listing on every search is far too slow on the hottest path. Maintain availability as its own authoritative, queryable calendar updated as bookings happen.

Keep a per-listing Availability calendar of open/taken nights. Search filters candidates by the requested range; booking will later reserve against the same calendar. One authoritative place for “is this home free?”.

What the new pieces do

Availabilitystore
Per-listing calendar of which nights are open or taken. Both search (filter by dates) and booking (reserve dates) depend on it being correct.

Step 4 · Never double-book

Transactional booking

Two guests try to book the same home for overlapping dates at the same instant. Without care, both succeed — and one shows up to a taken house. This is the marketplace’s cardinal sin.

API GatewayListings DBBooking ServiceAvailability
New in this step: Booking Service. · swipe to pan the diagram

Two guests try to book the same home for overlapping dates at the same instant. How do you guarantee only one wins?

  1. A gap between check and write is the classic race — both guests see "free", both write, both think they booked. Check-then-set must be one indivisible step.

  2. Confirming two overlapping stays means someone arrives at a taken house — the marketplace’s cardinal sin. The conflict must be prevented at write time, not reconciled after.

  3. Check the range is free and mark it taken together, guarded by a unique (listing, night) constraint or row locks. Concurrent conflicting bookings collapse to exactly one winner; the other cleanly retries.

The Booking Service reserves the requested nights in a single atomic transaction: check the range is free and mark it taken together, guarded by a constraint or lock on the listing+dates. Concurrent conflicting bookings — exactly one wins.

  • atomicreserve
  • 1winner per night
  • 0double-bookings

What the new pieces do

Booking Serviceservice
Turns “I want these dates” into a confirmed reservation — atomically, so two guests can’t book the same home for overlapping nights.

Back of the envelope

check + mark taken = 1 transaction
no gap for a race
unique (listing, night)
the DB enforces one winner
loser retries cleanly
0 double-bookings

Step 5 · Take the money

Payment hold & capture

Confirming a stay involves money across two parties, and payment can fail or be retried. Charging twice, or reserving dates without securing payment, both break trust.

Booking Servicereserve datesPaymentshold + capture
New in this step: Payments.

Confirming a stay involves money and the payment can fail or be retried. How do you take it safely?

  1. Reserving without securing funds means confirmed dates with no guaranteed payment, and no clean way to release them on failure. Tie the hold to the booking and release both if payment fails.

  2. Hold the card to secure funds while finalizing, capture on confirmation, and release the hold and the dates if it ultimately fails. An idempotency key makes the inherently retried payment path safe from double-charges.

  3. Charging before a booking is even confirmed means refunds, double-charges on retries, and money taken for stays that never happen. Separate auth from capture.

On booking, authorize (hold) the guest’s card; on confirmation, capture it and schedule the host payout. Use an idempotency key so a retried request never double-charges, and release the hold (and the dates) if payment ultimately fails.

What the new pieces do

Paymentsservice
Authorizes the guest’s card to hold funds at booking and captures on confirmation, with idempotency so a retry never double-charges.

Step 6 · Keep it fresh

Reviews & reindexing

New listings, price changes and reviews must show up in search, and ranking should reflect quality — but doing that work synchronously on the booking/listing path would slow everything down.

API GatewaySearch ServiceGeo IndexListings DBBooking ServiceAvailabilityPaymentsEvents
New in this step: Events. · swipe to pan the diagram

New listings, price changes and reviews must reach search, and ranking should reflect quality. How, without slowing writes?

  1. Doing reindexing and review aggregation inline makes every write wait on search-index updates, coupling a fast critical path to slow propagation. These effects are eventually-consistent by nature.

  2. Querying the source-of-truth DB per search reintroduces the read load on writes you separated in step 1, and can’t do fast ranking. Search needs its own index, kept fresh out-of-band.

  3. Listing/booking changes flow to an event stream; consumers reindex search, roll reviews into ratings, update payouts and notify — off the request path. Writes stay fast, search stays current (eventually).

Emit listing and booking changes as events. Consumers reindex the search service, aggregate reviews into ratings, update host payouts and send notifications — all asynchronously, so the write path stays fast and search stays current.

What the new pieces do

Eventsbus
A stream of listing and booking changes that drives search reindexing, review aggregation, host payouts and notifications asynchronously.

Back of the envelope

change ⇒ event
the write path doesn’t wait
consumers: reindex, reviews, payouts
eventually consistent by nature
record change ≠ propagate effects
decoupled via the stream

Step 7 · Scale the reads

The sharp edges

Search traffic dwarfs everything and spikes seasonally and by destination; popular cities create hotspots. One listings database can’t serve it all, and recomputing identical searches is wasteful.

GuestHostAPI GatewaySearch ServiceGeo IndexListings DBBooking ServiceAvailabilityPaymentsEvents
The system as it stands at this step. · swipe to pan the diagram

Add read replicas for the listings data, cache popular search results and listing pages, and shard/partition by region. Pricing and ranking can be precomputed; the rare booking path stays transactional while reads scale out horizontally.

You did it

You just designed Airbnb.

GuestHostAPI GatewaySearch ServiceGeo IndexListings DBBooking ServiceAvailabilityPaymentsEvents
The finished design, end to end. · swipe to pan the diagram

Everything you assembled, in order

  • A gateway splitting heavy search reads from rare, careful booking writes.
  • A spatial index (quadtree/geohash) makes “homes near here” a bounded query.
  • Per-listing availability calendars power both date filtering and booking.
  • Atomic, lock-guarded booking so overlapping stays can never double-book.
  • Payment hold-then-capture with idempotency keys for safe, retryable charges.
  • Event-driven reindexing and review aggregation keep search fresh asynchronously.
  • Read replicas, search caching and regional sharding scale the read-heavy load.

Where an interviewer pokes next

Getting the boxes right is the easy half. These are the questions that separate a candidate who drew the diagram from one who has run the thing. Answer each one out loud before you open it.

  1. How do you prevent double-booking under high concurrency?

    Make the reserve a single atomic operation: a unique constraint on (listing, night) so the second insert fails, or a row lock / SELECT … FOR UPDATE on the calendar range inside a transaction. Optimistic concurrency (version the calendar, retry on conflict) also works. The invariant is check-and-set as one indivisible step — never check, think, then write.

  2. Search shows a listing as available but it gets booked first — what happens?

    Expected: search is eventually-consistent and best-effort; the booking transaction is the authority. The guest picks dates that looked free, and the atomic reserve is the real check — if someone won the race, the booking fails and the guest picks again. You never rely on search availability being perfectly live.

  3. How do you keep the search index fresh as listings and bookings change?

    Emit change events from the write path; a consumer updates the search, geo and review indexes asynchronously. There’s a small lag, fine for discovery. You can push booking events at higher priority so date filters don’t go too stale, but the booking transaction still backstops correctness.

  4. How would you add surge / dynamic pricing?

    Compute it off the hot path: a pricing service consumes demand signals (searches, booking velocity, events, seasonality per region) and precomputes/caches a price per listing/date that search and the listing page read. Like ad-pricing or ride-hailing surge, it’s an async consumer feeding the read path, not a synchronous step in the booking transaction.

  5. Why optimize search and booking so differently?

    Opposite needs. Search is enormous, read-heavy, latency-sensitive and tolerant of slight staleness — scale it with replicas, caches, derived indexes, regional sharding. Booking is rare and must be exactly right — keep it transactional on the primary. Co-designing them would force one to compromise; separating lets each be optimal.

Check yourself — the answers, and why

Eight steps in, these are the calls you should be able to make cold. Pick one, then read why.

  1. Search runs against a derived index (not the write DB) so that…

    • It’s more accurate
    • Heavy read load never threatens the source of truth
    • Bookings are faster

    Separating search from the system of record lets reads scale without endangering writes.

  2. "Homes near here" is fast because of…

    • A lat/long B-tree
    • A spatial index (quadtree / geohash)
    • Sorting by distance

    Spatial indexing makes proximity a bounded viewport lookup, not a planet-wide scan.

  3. Double-booking is prevented by…

    • Checking availability before writing
    • One atomic reserve (unique constraint / lock on listing+nights)
    • Letting both through and refunding

    Check-and-set must be indivisible — the DB enforces exactly one winner per night.

  4. Payment uses authorize-then-capture so that…

    • It’s cheaper
    • You hold funds while finalizing and unwind cleanly on failure (idempotent)
    • Hosts get paid first

    Hold at booking, capture on confirm; an idempotency key makes the retried path safe.

  5. Reindexing and review aggregation run asynchronously because they’re…

    • Unimportant
    • Eventually-consistent — kept off the fast write path
    • Done by the client

    Emit change events; consumers propagate effects so writes stay fast and search stays current.

How you’d open this design in an interview

Before any boxes: agree what it must do, pin the qualities that shape everything, then build — naming each trade-off as you make it. The walkthrough above is that exact order.

What it must do

Agree on these before drawing a single box.

  • List a home: a host publishes a listing with photos, price and an availability calendar.
  • Search: guests query by place, date range, price and amenities, ranked for relevance — the hot path.
  • Availability: results exclude homes already taken for the requested nights.
  • Book: reserve a contiguous range of nights atomically — never double-book a home.
  • Pay & review: hold-and-capture the guest’s card on confirmation; reviews feed ranking.

The qualities that shape everything

Each one names the mechanism that buys it.

Heavy search never threatens the source of truth
Search reads a derived index while host writes go to the Listings DB — so a search overload or a bad reindex never endangers the booking-critical data.
“Homes near here” without a planet-wide scan
A spatial index (quadtree / geohash) buckets listings by region, turning proximity into a bounded viewport lookup over a small candidate set.
Filter by date range
A per-listing availability calendar of open/taken nights that both search filtering and the booking reserve read from.
Never double-book a home
The booking service reserves the nights in one atomic transaction, guarded by a unique (listing, night) constraint or row lock — concurrent conflicts collapse to a single winner.
Safe, retryable payments
Authorize (hold) at booking and capture on confirm, with an idempotency key so a retried charge never double-charges and the dates release if payment fails.
Search stays fresh without slowing writes
Listing and booking changes emit events; consumers reindex search and aggregate reviews asynchronously, off the write path.
Absorb read-heavy, spiky traffic
Read replicas for listings, cached popular searches and listing pages, and regional sharding — while the rare booking path stays transactional on the primary.

The trade-offs you say out loud

Senior signal isn’t the boxes — it’s naming what you gave up and why it was the right price.

A derived search index over searching the write database directly

Running heavy search straight against the source-of-truth DB makes reads contend with listings and bookings and can’t scale on its own. A separate index lets search scale, and a search overload never threatens the writes.

A spatial index over a lat/long B-tree

A 1-D index can’t answer 2-D proximity — nearby on one axis can be far on the other, so you still scan huge ranges. A quadtree/geohash makes “near here” a bounded region lookup.

One atomic reserve over check availability then write

A gap between check and write is the classic race: both guests see “free”, both write, someone arrives at a taken house. A unique (listing, night) constraint makes check-and-set indivisible.

Authorize-then-capture over charging immediately

Charging before the booking is confirmed means refunds, double-charges on retries and money taken for stays that never happen. Hold-then-capture unwinds cleanly, and an idempotency key makes the retried path safe.

Async reindex via events over synchronous reindex on the write

Reindexing and review aggregation inline make every write wait on slow index updates. These effects are eventually-consistent by nature, so an event stream keeps writes fast and search current.

What this teaches

Learn system design by building a lodging marketplace like Airbnb step by step. An interactive guide covering listings, geo search with filters, availability calendars, booking without double-booking, the payment hold-and-capture flow, reviews and reindexing, and scaling reads.

Key takeaways

  • A gateway splitting heavy search reads from rare, careful booking writes.
  • A spatial index (quadtree/geohash) makes “homes near here” a bounded query.
  • Per-listing availability calendars power both date filtering and booking.
  • Atomic, lock-guarded booking so overlapping stays can never double-book.
  • Payment hold-then-capture with idempotency keys for safe, retryable charges.
  • Event-driven reindexing and review aggregation keep search fresh asynchronously.
  • Read replicas, search caching and regional sharding scale the read-heavy load.

Concepts covered

  • What is Airbnb?
  • List and find homes
  • Geo indexing
  • Availability calendars
  • Transactional booking
  • Payment hold & capture
  • Reviews & reindexing
  • The sharp edges
built to be booked, not memorized — make the calls, drop the search, run the gauntlet.
Finished this one? 0 / 65 System Designs done

Explore the topic

See this alongside everything else on the same subject — handbooks, system designs, challenges and tools, in one place.

More System Designs