Original data study
We scanned 35 London hotels across 5 OTA channels on 4 October 2026. 94% of them had at least one factual field where their own listings disagreed, and the average consistency score was 81.9 out of 100.
Method and page reviewed 3 October 2026 · Scan window 4 October 2026 · full method and limits
94%
Properties with at least one mismatch
33 of 35 scored properties
81.9
Average consistency score
Median 86.0 · 0–100 scale
35
Properties scored
3 not measured (error)
15.4
Fields compared per property
Average count of fields with comparable evidence
A "mismatch" is a field where at least two successfully read listings stated comparable values that do not agree after normalisation. A field no listing stated, or that only one listing stated, is reported as not comparable and is excluded from every rate on this page.
Of the properties where a field could actually be compared, the share with at least one conflicting value in that group.
| Field | Properties with a conflict | Conflicting of comparable |
|---|---|---|
| Breakfast | 17% | 6 of 35 |
| Check-in/out | 33% | 11 of 33 |
| Pet policy | 37% | 13 of 35 |
| Parking | 68% | 23 of 34 |
| Cancellation | 0% | 0 of 30 |
| Fees & taxes | 64% | 9 of 14 |
| Wi-Fi | 9% | 3 of 35 |
| Children | 59% | 20 of 34 |
| Amenities | 9% | 3 of 35 |
For each channel, the share of the field values it supplied that sat inside a disagreement. This does not say the channel is wrong — a conflict involves at least two listings, and this counts participation in one, not fault.
| OTA | Values in a conflict | Conflicting of supplied |
|---|---|---|
| Booking.com | 20% | 85 of 430 |
| Expedia | 19% | 90 of 466 |
| Hotels.com | 18% | 77 of 438 |
| Trip.com | 0% | 0 of 2 |
| Agoda | 18% | 68 of 383 |
A rate above is only as good as the evidence behind it. Some channels serve a bot-verification page instead of the listing, and a channel we could not read contributes nothing — it is never counted as agreement, and a low conflict rate on a channel with few readable listings says more about our access than about that channel's accuracy.
| OTA | Listings found | Read successfully | Could not read |
|---|---|---|---|
| Booking.com | 35 | 35 | 0 |
| Expedia | 35 | 31 | 4 |
| Hotels.com | 34 | 30 | 4 |
| Trip.com | 35 | 3 | 32 |
| Agoda | 35 | 34 | 1 |
Real conflicts from the sample, with the property described rather than named. Values are quoted as the listings stated them.
Five-star hotel, central London
Two Expedia Group brands and Agoda describe the same stay differently. A guest booking on Agoda expects breakfast; the hotel expects to charge for it. This is the mismatch most likely to be discovered at the table on the first morning.
Five-star hotel, central London
The largest single money gap in the sample. A family booking through either Expedia Group brand has been told an extra bed is free; through Booking.com, that it costs GBP 100 per person per night. Both cannot be right, and the difference lands at check-in.
Five-star hotel, central London
A guest who books on Expedia and arrives by car has been told parking exists. Booking.com says it does not. Whichever is correct, one set of arrivals is being set up for a bad first five minutes.
| Location | London, United Kingdom |
|---|---|
| Properties attempted | 38 |
| Properties scored | 35 |
| Read but not comparable (fewer than two readable sources) | 0 |
| Not measured (discovery or request error) | 3 |
| Listings read successfully | 133 |
| Channels read | Booking.com, Expedia, Hotels.com, Trip.com, Agoda |
| Scan window | 4 October 2026 |
| Run identifier | 2026-10-04T05-30-58-115Z |
Every OpenStreetMap feature inside the Greater London administrative boundary tagged tourism=hotel with stars 4 or 5, as of the OSM snapshot above. Unnamed features removed; duplicates removed by name and by position rounded to 4 decimal places; the remainder ordered by OSM element id and every 5th taken, up to 38.
The star rating used to select these properties is the one recorded in OpenStreetMap, which is crowd-sourced. It is not an AA or VisitEngland accreditation, and nothing here is a claim about any property's official rating — it defines which properties were scanned, and nothing else.
The six steps every study follows, how a mismatch is defined, and the eight limits that apply to every rate on this page are set out once on the consistency study hub. They are not restated per city, so there is one authoritative copy to check them against.
Suggested citation
MisMatchMaker AI, "Hotel OTA listing consistency benchmark: London", 3 October 2026. https://mismatchmaker.com/hotel-listing-consistency-benchmark/london
The aggregate figures on this page are published under the Creative Commons Attribution 4.0 licence: reuse them anywhere, including commercially, with attribution and a link back to this page. If you quote a rate, quote its sample size and scan window with it — a share without a denominator is not a finding. Raw per-property reports are not published, because they contain identifiable properties and third-party listing text.
40 properties scanned 5 October 2026 – 6 October 2026, with 65% showing at least one mismatch.
Every four- and five-star hotel in Amsterdam scanned across five channels: how many disagreed with themselves, on which fields, and on which channel.
Where breakfast, check-in, pets, parking and cancellation are set in the Extranet, each path quoted from Booking.com's own help centre.
Where each channel draws the child age line, whether a free child counts against occupancy, and the family search test that shows what guests are quoted.
Score your own process across five dimensions in two minutes and get a maturity baseline you can retest later.
This study runs the same free check anyone can run. Scan your own property to see which of these fields disagree across your listings, with the evidence from each channel beside it.
Run a free match check