What hotel and travel data is most valuable and hardest for Expedia, Booking, Google and AI platforms to collect themselves? What do travel companies currently pay or license for hotel content, photos, reviews, mapping and performance data — named companies and price or deal benchmarks (e.g. CoStar/STR, Lighthouse, Amadeus, Hotelbeds, DerbySoft)? And how would verified, in-person, time-stamped, room-level hotel data plus the questions travelers ask compound into a defensible data moat those companies would want to license?
Evidence base: 140 source-linked records gathered to 2026-10-24 — 60 public commercial benchmarks (acquisitions, funding, product and API prices, coverage metrics), 35 transaction and funding records with named buyers, and 45 documented examples of what hotel data fields incumbents can and cannot see. Amounts keep their source currency and date; whole-business deal values are not like-for-like dataset license prices.
Left: strategic deal and funding benchmarks (log scale, source currency, deduplicated by deal) against public product starting prices — the gap between a $100/month rate feed and a $1.52B data-and-distribution acquisition is four orders of magnitude. Right: the collection flywheel in which traveler questions act as a demand oracle that schedules the next in-person audit. Hover any bar or step for detail.
Qualitative scoring (1–5 dots) of value to licensees × collection difficulty × refresh burden × rights defensibility, synthesized from the 45 field-level evidence records. Incumbent APIs expose names, rates, amenity flags and reviews; none return measured room geometry, time-of-day noise, in-room Wi-Fi, accessibility clearances or verified amenity operating status.
Eight linked layers, joined at canonical hotel → building/tower → room category → sampled physical room, every observation carrying observed_at, valid_from/valid_to, evidence hash, confidence and rights scope.
Canonical property ID; brand, owner, operator; address, lat/lon, entrance coordinates; tower/building, floor plan; room-category aliases by Expedia, Booking, Google and the hotel; supplier IDs and merge/split history.
Marketed category vs sampled physical room (number tokenized), floor, measured square footage, layout, bed type/count/measurements, occupancy, connecting-door truth, bathroom type, shower/tub, grab bars and clearances, outlets/USB, desk, HVAC, window openability, blackout, safe/laptop fit, view direction and obstructions, elevator distance and accessible route.
Cleanliness, odor/smoke, wear, mattress and linen condition, water pressure and temperature, light and sound levels by time of day, noise sources (corridor, street, club, elevator), in-room Wi-Fi SSID/speed/latency, cellular signal, housekeeping frequency, pest/maintenance observations, renovation state.
Breakfast contents, hours and crowding; pool/gym/spa status and hours; parking price, clearance, EV chargers; shuttle, lounge, kids/pet rules, beach/ski access, laundry, restaurant and room-service hours, accessibility equipment, construction, closures and outages.
Hotel/building/room/category IDs, captured_at, GPS, compass direction, device and collector, original EXIF and hash, scene label, before/after renovation link, manipulation record, ownership and license scope, releases, expiration and takedown status.
Room/rate-plan, observed_at, stay dates, price, taxes/fees/resort fee, inclusions, refund/cancel/payment terms, availability, parity, minimum stay, sellout; occupancy, ADR, RevPAR, forward demand, profitability where licensed; conversion, cancellation, refund and support-contact outcomes.
Raw question, normalized intent, segment and trip context, hotel/building/room target, asked_at, frequency, answer, evidence IDs, answered_at, confidence and freshness, contradiction/unresolved status, helpfulness, booking and post-stay validation.
Observation ID, collector and method, source class, observed_at, valid_from/valid_to, location accuracy, units, evidence hash/URL, confidence, QA reviewer, rights and permitted uses.
Every pricing, deal and scale record behind the charts, labeled by kind so unlike units are not read as comparable. Click a header to sort. Major deals appear under several corroborating sources; read them as one deal each.
| Provider ↕ | Kind ↕ | Amount / metric ↕ | Scope | Source |
|---|
Method: 140 records from three result sets researched to 2026-10-24 — 60 public commercial benchmarks, 35 transaction/funding records, 45 field-level evidence examples — each linked to its source page. Chart deduplicates corroborating deal reports to one bar per transaction and keeps source currencies; deal values price whole businesses (technology, customers, distribution), not datasets alone, and many enterprise license terms are quote-based and unobservable. Scorecard dots are qualitative synthesis, not source numbers. Some third-party-reported prices (e.g. the Google Place Details benchmark) are as cited by the reporting site. Evidence excerpts trimmed for space.