Ever wonder what information powers AI travel recommendations? In short: online reviews, YouTube videos, and recent, date-stamped guides. According to 5WPR’s new analysis of 632,542 Google AI answers, review titan TripAdvisor appears in 7.8% of "things to do" responses, while official destination marketing organizations (DMOs) are near absent.
Today, 56% of U.S. leisure travelers rely on AI to research and plan trips, so knowing where these models get their information is crucial for any travel brand.
Read on to discover the sources shaping AI recommendations and how you can get cited.
What are the key findings of the 5WPR new travel research?
The study reveals AI models select travel sources based on question type: hotel queries rely on review platforms, hidden-gem queries retrieve creator videos and community posts, and itinerary queries extract data from dated, practical guides.
The seven findings below detail the specific metrics and sample sizes behind our research.
| Finding | Number | Sample |
|---|---|---|
| Tripadvisor is the source Google AI cites most for "things to do in" questions. | 7.8% of answers | 632,542 Google AI answers |
| Video and social platforms (YouTube, Facebook, Instagram, TikTok) appear in Google's things-to-do answers. | 15.1% of answers | 632,542 Google AI answers |
| Tourism board sites were absent from Google's top ten sources; Visit Florida and the National Park Service make ChatGPT's. | 2.0% of ChatGPT answers | 7,178 ChatGPT answers |
| Review and booking sites (Tripadvisor, Google, Expedia, Booking.com, Hotels.com) supply Google's hotel answers. | 30.7% of answers | 12,009 Google AI answers |
| YouTube is the top source for "hidden gems" questions, and Tripadvisor drops to third. | 9.4% vs 6.6% | 2,527 Google AI answers |
| Each boutique hotel ChatGPT named in Charleston, South Carolina, was rated 4.6 or higher. | Average 4.87 of 5 | 7 hotels, one live ChatGPT run |
| Every text source ChatGPT cited for Savannah, Georgia, was a tourism guide site, and every dated page was under seven weeks old. | 4 of 4 citations | One live ChatGPT run |
Source: 5WPR Travel AI Citation Study, November 5, 2026, built on DataForSEO LLM Mentions data and live ChatGPT queries. Shares overlap because single AI answers cite multiple websites.
Which websites does Google AI cite for things-to-do travel queries?
Google AI cites Tripadvisor, YouTube, and Facebook most frequently for "things to do in" questions in the United States. 5WPR analyzed 632,542 Google AI answers using the Tripadvisor surfaced in 7.8% of Google AI answers, followed by YouTube at 5.5%, Facebook at 5.1%, Reddit at 3.9%, and Instagram at 3.2%.
Video and social platforms (YouTube, Facebook, Instagram, and TikTok) generate 15.1% of Google AI citations. Booking and event platforms complete the top sources: Eventbrite captures 2.0%, Expedia earns 1.5%, and GetYourGuide secures 1.2%. Single Google AI answers cite multiple websites, preventing the shares from totaling 100 percent. Destination marketing organizations are largely absent in the current Google AI top ten sources.
How do ChatGPT travel recommendation sources differ from Google AI sources?
ChatGPT cites editorial travel guides and official tourism websites at higher rates than Google AI. Across 7,178 ChatGPT answers to identical "things to do in" queries, Tripadvisor maintained the top position at 6.8%. Reddit followed Tripadvisor at 2.5%, and Lonely Planet ranked third at 2.1%.
Three editorial travel guides (Lonely Planet, Atlas Obscura, and Time Out) captured 4.6% of ChatGPT citations. Two official sources, the National Park Service and Visit Florida, generated 2.0% of ChatGPT citations. Google AI excludes both editorial guides and official sources from the top ten citation list.
| Rank | Google AI source | Share of answers | ChatGPT source | Share of answers |
|---|---|---|---|---|
| 1 | Tripadvisor | 7.8% | Tripadvisor | 6.8% |
| 2 | YouTube | 5.5% | 2.5% | |
| 3 | 5.1% | Lonely Planet | 2.1% | |
| 4 | 3.9% | 1.5% | ||
| 5 | 3.2% | Atlas Obscura | 1.5% | |
| 6 | Eventbrite | 2.0% | Wikipedia | 1.5% |
| 7 | Expedia | 1.5% | National Park Service | 1.1% |
| 8 | TikTok | 1.2% | Yelp | 1.1% |
| 9 | GetYourGuide | 1.2% | Time Out | 1.0% |
| 10 | Yelp | 1.1% | Visit Florida | 0.9% |
Source: 5WPR Travel AI Citation Study, analysis of DataForSEO LLM Mentions data, U.S. English questions containing "things to do in," pulled November 5, 2026. Google sample: 632,542 answers. ChatGPT sample: 7,178 answers.
Official tourism board pages earn direct citations within ChatGPT answers. To gain visibility on Google AI, tourism boards should publish content on owned sources and earn third-party mentions in places like Tripadvisor listings, YouTube channels, and Facebook pages. The October 2026 5WPR Airlines & Hotels AI Visibility Index identified an identical platform split: community and social websites generated 26.5% of Google AI Overview citations, compared to 3.0% across ChatGPT, Claude, Gemini, and Perplexity.
How do guest reviews influence AI hotel recommendations?
Review and booking platforms provide 30.7% of the citations powering Google AI hotel recommendations. 5WPR analyzed 12,009 Google AI answers to "best hotels in" queries. Tripadvisor generated 10.0% of citations, Google Business Profiles produced 8.8%, Expedia earned 4.9%, and Booking.com and Hotels.com captured 3.5%. Condé Nast Traveler and U.S. News Travel generated 5.0% of Google AI hotel citations.
When 5WPR prompted ChatGPT for the best boutique hotels in Charleston, South Carolina, on November 5, 2026, the model recommended seven properties. All seven boutique hotels maintained guest ratings between 4.6 and 5.0, averaging 4.87. ChatGPT displayed the numerical rating next to the hotel names. Review volumes ranged from 361 reviews for The Pinch to 4,547 reviews for the French Quarter Inn.
ChatGPT cited four text sources for the Charleston hotel recommendations: the Charleston Area Convention and Visitors Bureau historic inn guide and three Condé Nast Traveler articles. One Condé Nast Traveler article featured the 2026 Readers' Choice list. A second Condé Nast article featured a local hotel guide written by a guest who stayed at the property. The Pinch secured a recommendation with only 361 reviews, showing that review volume is just one of multiple factors in AI recommendations.
Which websites do LLMs cite for hidden gems travel queries?
For "hidden gems" queries, Google AI prioritizes creator videos, community discussion posts, and local guides, dropping Tripadvisor to the third most-cited source. Across 2,527 Google AI answers to "hidden gems in" questions, YouTube generated 9.4% of citations, Reddit earned 7.1%, Tripadvisor captured 6.6%, and Facebook supplied 6.4%.
Atlas Obscura, a database cataloging unusual destinations, generated 4.7% of hidden-gem citations after failing to reach the top ten for general things-to-do queries. Creator and community platforms (YouTube, Reddit, Facebook, Instagram, Quora, and TikTok) produced 29.5% of Google AI hidden-gem citations.
Google AI processes video citations differently than the text-first AI models evaluated in the April 2026 5WPR Creators & AI Visibility audit. The April 2026 audit found zero YouTube videos across 109 citations, noting 78% of travel citations originated from specialist publications. For Google AI hidden-gems questions, YouTube ranks as the highest-cited platform.
Travelers rely on ChatGPT and other AI models primarily for discovery and increasingly with planning. The July 2026 Longwoods International traveler survey revealed 54% of travelers using AI trip planning used the tools to discover attractions and activities. Longwoods International President and CEO Amir Eylon designated trip planning as an ideal artificial intelligence application. Earning AI visibility for lesser-known neighborhoods, trails, or restaurants requires securing coverage from local residents, video creators, and specialist guidebooks.
Which sources did ChatGPT cite for Savannah, Georgia travel recommendations?
When prompting ChatGPT for the best things to do in Savannah, Georgia, it retrieved ten attractions from two tourism guide websites as sources, and both were under seven weeks old.
ChatGPT cited Explore Georgia, the official state travel portal, and savannahtourism.org, an independent visitor guide. Publishers posted the three savannahtourism.org pages between September 23, 2026, and October 20, 2026. Three of the four cited web page titles contained the year "2026". One page title explicitly offered "40 Picks, With Hours and Prices".
Visit Savannah, the official destination marketing organization, provided 8 of the 32 images in ChatGPT answers, while Explore Georgia provided 2.
What content should travel businesses publish to earn AI travel recommendations?
The new 5WPR research pairs business types with proven data sources, serving as a checklist for travel public relations strategies.
| Business type | Source | Publication Recommendations |
|---|---|---|
| Tourism board (DMO) | ChatGPT: official sites and editorial guides. Google: Tripadvisor, YouTube, Facebook. | "Things to do" pages with the current year, hours and prices; coverage in Lonely Planet, Time Out and Atlas Obscura; video walk-throughs on YouTube. |
| Hotel | Tripadvisor, Google listings, Expedia, Booking.com, Hotels.com, Condé Nast Traveler. | A steady flow of new reviews with replies; matching name, address and photos on booking sites; pitches for editorial lists and readers' awards. |
| Tour or attraction operator | Tripadvisor, GetYourGuide, Eventbrite, Expedia. | Full listings with duration, price and meeting point; dated event pages; a review request after tours. |
| Local business or lesser-known site | YouTube, Reddit, Atlas Obscura, Facebook, Instagram. | Creator visits, an Atlas Obscura entry, answers in city subreddits from a disclosed account, a reviewed Google Business Profile. |
Travel businesses publishing exclusively on owned domains disappear from AI recommendations. In the hotel sector, third-party review and booking sites generate 30.7% of Google AI citations. 5WPR travel programs combine earned media coverage with hospitality digital marketing to populate owned and third-party platforms.
Structured, dated facts transfer effectively across LLM citation sources. The 5WPR Harvest Hosts AleBlazer campaign converted a 500-location milestone into 84 media placements. Media placements shared similar information, allowing traditional search engines and AI models to cross-verify and retrieve the data at the same time.
How did 5WPR collect the data in its new research?
This new research used two primary sources on November 5, 2026. The first data source utilized the DataForSEO LLM Mentions database, filtering U.S. English questions containing "things to do in," "best hotels in," or "hidden gems in," and grouping the results by the cited domains. The second data source incorporated two live ChatGPT web search queries analyzing Savannah attractions and Charleston boutique hotels.
Each percentage share represents the fraction of total AI answers citing a specific website. Because individual AI answers cite multiple websites, the percentage shares overlap. To cite the study data, use: 5WPR, Travel AI Citation Study, November 5, 2026.
What do AI travel recommendation sources have in common?
AI models retrieve travel information from trusted sources in their training data and web search capabilities, third-party mentions, user reviews, and brand websites. Google’s traditional search model and AI mode extract hotel answers from numerical ratings, hidden-gem answers from UGC, and itinerary answers from dated logistical guides. Destinations that maintain active profiles across multiple formats can increase their chances of appearing in AI recommendations.
How can 5WPR help travel brands increase their recommendations in AI answers?
The 5WPR Destination Source Map shares how AI travel recommendations originate from reviews, local guides, and updated destination pages. 5WPR partners with tourism boards, hotels, and operators to build modern digital footprints across earned and owned media sources to support machine retrieval and recommendations.
Founded in 2003, 5W has continued to evolve its approach to communications, bringing together public relations, digital marketing, social media, influencer partnerships, reputation management, and AI visibility. As the ways audiences discover, evaluate, and engage with brands change, 5W helps clients build visibility and credibility across earned media, search engines, and AI-powered platforms, including ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews. Start a conversation with 5W today.
