Sigtrip.

AI Hotel Ranking · New York City

When travelers ask ChatGPT where to stay in New York, which hotels does it name?

A standing census of what AI actually recommends — the hotels ChatGPT surfaces for real guest questions, ranked overall and by neighborhood. Not our opinion: the engine’s own answers.

657+
NYC hotels tracked
87
guest questions asked
242
hotels ChatGPT named
1
engine (ChatGPT)

Updated August 6, 2026 · ChatGPT with web search

Is your hotel on the board — or invisible?

Enter your website to see exactly where ChatGPT places you across every guest query — and start improving it. Free ChatGPT scan, no card.

The ranking

Pick a board and a metric. Percentages are how often ChatGPT names a hotel across the relevant questions — placement, not a score we assign.

Rank by

Overall — all NYC · 55 hotels ChatGPT names · based on 88 answers to guest-style questions. Placement rate = share of those answers that name the hotel.

#HotelPlacement rate#1 rate
1The Greenwich HotelTribeca54.5%8%
2The Whitby HotelMidtown West31.8%2.3%
3Nine OrchardLower East Side27.3%2.3%
4Aman New YorkMidtown / Fifth Avenue26.1%10.2%
5The LowellUpper East Side25%0%
6The St. Regis New YorkMidtown East21.6%0%
7Waldorf Astoria New YorkMidtown East20.5%0%
8The Algonquin Hotel Times Square, Autograph CollectionMidtown18.2%0%
9The KnickerbockerTimes Square15.9%0%
10The Pierre, A Taj Hotel, New YorkUpper East Side15.9%0%
11The Ritz-Carlton New York, NoMadNoMad14.8%2.3%
12The Standard, High LineMeatpacking District14.8%0%
13The Manner, The Unbound Collection By Hyatt Hotel New York13.6%1.1%
14Four Seasons Hotel New York DowntownTribeca12.5%0%
15The Ludlow HotelLower East Side12.5%0%
16Warren Street HotelTribeca12.5%0%
17Wythe HotelWilliamsburg12.5%0%
18The William ValeWilliamsburg11.4%0%
19Mandarin Oriental, New YorkColumbus Circle10.2%0%
20The Hoxton, WilliamsburgWilliamsburg10.2%0%

Where ChatGPT looks

Every recommendation is built from sources the engine cites. Here's the mix behind the New York answers — and how often it reaches a hotel's own website versus an OTA.

What kind of source ChatGPT cites

  • Hotel's own site
    59.7%
  • Guides & listicles
    17.4%
  • Other
    7.1%
  • Media & editorial
    6.1%
  • Review sites
    4.1%
  • OTAs (Booking, Expedia…)
    2.2%
  • Transit & airports
    2%
  • Reference (Wikipedia…)
    1.2%

Most-cited sources

  • marriott.com10.2%
  • hyatt.com7.1%
  • Time Out6.5%
  • hilton.com6.1%
  • MICHELIN Guide4.7%
  • Condé Nast Traveler3.9%
  • Tripadvisor3.9%
  • ihg.com2.4%
  • Forbes Travel Guide1.9%
  • NYC Tourism1.4%
  • 1hotels.com1.2%
  • fourseasons.com1%

Questions

What does this ranking actually measure?

It measures the AI engine, not the hotel. We ask ChatGPT the questions real travelers ask — 'best hotels in NYC', 'where to stay in Williamsburg' — and record which hotels it names and how it ranks them. A hotel's placement rate is the share of those answers that name it.

Why percentages instead of raw numbers?

A placement rate — 'named in 41% of relevant answers' — is comparable across boards and honest about coverage. We publish the percentage and the ordinal position, and keep the underlying sample sizes and confidence intervals on the methodology page.

How is a neighborhood board different from the overall board?

The overall board reflects citywide questions ('best hotels in New York'). A neighborhood board reflects the questions for that area ('hotels in Williamsburg'). A hotel appears on a board because ChatGPT names it for those questions — not because of how we tagged it.

Which AI engine is this?

ChatGPT with web search, today. Each engine answers differently, so we never blend engines into one score — other engines will get their own boards as we add them.

My hotel isn't listed. What does that mean?

It means ChatGPT didn't name your property for these questions in this run — the most common reason a hotel loses AI-driven bookings. Run a free scan to see exactly where you stand and why.

How often does it update?

The census re-runs on a regular cadence; every board is stamped with the run it came from. Figures are a snapshot of what the engine said at that time, not a permanent verdict.

Do you rank your own customers?

No. Sigtrip customers are recorded and excluded from placement, and a set of hotels is held out entirely to validate the method. See the methodology page for the full disclosure.

How we measure this — the questions, the sample sizes, the confidence intervals, and what we exclude — is fully documented.

MethodologyResearch & findings