Hotel technology & trends

Your Website Didn't Die. It Changed Jobs.

By Diego Verzini · 4 min read

Sixteen experts were asked this week whether the hotel website has become the least important part of the digital journey. Almost none said yes. Almost all said the same thing in different words: it has a different job now.

For fifteen years the website was the shop window, and the redesign was the project of the year. I've signed off on three of them. The data says the next one should probably not happen.

Discovery moved out

  • Hotel websites get about 6% of the citations in AI travel answers — and nearly all of that is two big chains. OTAs and metasearch get 44%. Reddit alone gets 17%, the single most-cited domain (cloro, 1,548 answers across six engines).
  • 94.3% of accommodation websites were never cited by any AI model. 1,337 sites, 450 queries, four models, five European countries. Only 76 properties showed up at all (Nokumo, 2026).
  • Google counts 48 touchpoints in a travel journey. The hotel's site is one of them, and it's rarely the first.

The guest doesn't find you on your website anymore. She finds you on the OTA, on Maps, in a chat. Then she comes to your site to check one thing.

What the machines actually read

Here's the part that changes the budget. Everyone else is now reading your site to describe you.

  • AI cites hotels on narrow questions, not on "best hotel in town." Parking, pets, late check-in, cancellation policy. The page that wins is the boring one with the exact answer.
  • AI repeats your mistakes; it doesn't invent them. An audit found a West Hollywood hotel described with Nashville details and a resort promoted with a restaurant that closed years ago. Hotel and brand sites appear in 31.6% of ChatGPT answers — so what's on them matters.
  • The OTA reads it too — out loud. On October 29 Expedia launched "Property Expert": AI that answers a traveler's questions about your hotel, before and after booking, from your listing, your reviews and their reservation. If your listing says the wrong parking hours, the OTA now says them with its own voice. Not an enemy — the most attentive reader of your listing.

Where the money goes instead

WorldHotels — 4,000 independent hotels under one roof — just rebuilt its site. Not for a new look: it spent 10,000 hours interviewing franchisees for the local facts that feed its channels and the AI platforms. Its CTO's reading of the road ahead: today the assistant discovers and bounces you to the site to book; next, the transaction happens inside the chat; after that, the guest's agent talks to the hotel's agent.

The experts' consensus follows the same logic. Stop the cosmetic redesign. Move the money to:

  • Policy pages that answer exactly. One page, one question, one answer — the 50-question sheet we wrote about in September, now published where machines read it.
  • The handoff to the booking engine. Visitors arriving from AI reach the engine more often than average visitors. Don't lose them in a second redirect.
  • Measurement. GA4 only catches AI visits that carry a referrer; AI Overview traffic hides inside organic. Reconcile the booking engine with the PMS, or you'll never know what the chat sent you.

One more number from the panel: AI has cut the effort of a redesign by more than half. The site got cheaper to build precisely when building it stopped being the point.

The takeaway: your website's job is no longer to be found. It's to be the source of truth — for the guest who comes to check one detail, for the OTA that answers in your name, for the AI that describes you to a stranger. Three moves for a 30-room hotel: (1) cancel the redesign and spend that money on exact policy pages and a clean pass to the booking engine; (2) the same 50 answers on your site, in your engine and on every OTA listing, because the OTA now reads them aloud; (3) once a month, ask a chat about your hotel and fix the site where the machine got it wrong. The machine keeps the 50 answers identical everywhere, 24/7. You decide what the answers are.

See you at check-in,
Diego
Sources
Hospitality Net — World Panel "Have hotel websites become the least important part of the digital customer journey?" (October 29, 2026): 16 expert responses, including Max Starkov (Google's 48 touchpoints; cloro and Nokumo data), Alessio Re of Klartika (AI Overviews cite hotels on narrow questions such as parking, pets, late check-in and cancellation; GA4 and Search Console measurement gaps), Lynn Patchett of Kollective (AI visitors reach the booking engine more often than average), Christine Malfair (AI has cut redesign effort by more than half).
cloro — "Hotel Travel SEO: Which Hotels ChatGPT Recommends" (June 2, 2026, updated September 15, 2026): 1,548 AI answers across Google AI Mode, ChatGPT, Google AI Overview, Gemini, Copilot and Perplexity — hotel websites about 6% of citations (Marriott 3%, Four Seasons 3%), OTAs and metasearch about 44%, Reddit 17%.
Nokumo Research — "AI Visibility in Hospitality: What 3,600 AI Responses and 1,337 Website Audits Reveal" (March 2026): 450 queries, four models, Germany, Austria, Switzerland, Slovenia and Croatia — 5.7% of audited properties detected in any AI response (94.3% never cited).
Martin Soler — Tell #177, "Who's actually asking ChatGPT?" (October 29, 2026): AGR AI referral audit — AI recommendations repeat website errors; hotel or brand sites cited in 31.6% of ChatGPT answers.
PhocusWire — "BWH rebrands to WorldHotels, unifies loyalty programs, boosts digital offering" (October 27, 2026): 4,000 hotels, 18 brands; website and app rebuilt; 10,000 hours of franchisee interviews for content feeding channels and AI platforms; rates and availability made accessible to AI engines; CTO Bill Ryan on discovery in the assistant, transaction in the chat, agent-to-agent.
Expedia Group — Fall 2026 Product Release (October 29, 2026, via Hospitality Net and PhocusWire): "Property Expert" answers property questions before and after booking from property details, guest reviews and the traveler's reservation; "AI Compare" on Hotels.com compares properties on location, amenities, price and policies.
← Back to the blog