The change almost no Swiss hotelier notices

If you googled "Hotel Lugano lakeside" in 2024, you saw 10 blue links plus a map. In 2026, you see first an AI-generated answer that recommends three hotels, each with two-sentence reasoning and star count. The blue links come after — if you scroll.

The numbers behind it: according to current studies (Innsight 2026, Altos Agency 2026), 68% of local searches now show AI Overviews. For hotels that means: if your house isn't among the three AI-selected recommendations, you see fewer clicks — not just fewer clicks from position 4, but often no clicks at all.

Organic click-through rates in local hotel searches dropped 30 to 45% in many markets in 2026, according to hospitality marketing studies. Swiss hoteliers feel this without knowing the cause — the profile looks good, reviews are stable, but the phone rings less.

This piece is the explanation. And the four levers that still work.

What changed technically

Briefly and without SEO jargon: Google's AI — the one behind AI Overviews — learns from three data sources whether a hotel is "recommendable":

Structured data on the hotel website (Schema.org markup, FAQ sections, clear Q&A structures)
Google Business Profile (star count, review frequency, reply consistency, photo recency)
External mentions (in blogs, travel guides, AI-trained datasets)

The old lever: keywords on the website. The new lever: clear, structured answers to typical travel questions. AI models read FAQ sections, tables, well-organised texts better than rambling marketing prose. And they quote what they can read.

Lever 1 — Structured answers on your own website

The most important new mechanic: FAQ sections on the hotel website that answer typical travel questions. Not "Why our hotel convinces" — but specifically:

"How far is the train station?"
"Which rooms have a lake view?"
"Is there a complimentary shuttle to the ski area?"
"Is wellness included in the room price?"
"What's the cancellation policy?"
"Which restaurants do you recommend nearby?"

These question-answer pairs need to be tagged with JSON-LD schema (FAQPage schema) so the AI recognises them as citable answers. Anyone with this gets listed in AI Overviews as a recommendation, often with a direct quote from the website's own text.

Concretely: 15 to 20 FAQ answers, 2 to 4 sentences each, with a clear answer structure. A weekend project for a location manager, not a marketing budget.

Lever 2 — Review frequency as an AI signal

AI models like Google Gemini behind AI Overviews weight "fresh" reviews more heavily than old ones. A hotel with 200 reviews all from 2024 looks less relevant to the AI than a hotel with 80 reviews, 20 of which are from the last 60 days.

That has consequences for review strategy: consistent frequency beats large one-off spikes. 4 to 6 new reviews per month, evenly distributed, beats 30 reviews in April and nothing afterwards.

The same logic applies to reply frequency: anyone replying to old reviews only now signals to the AI "hotel is currently active" — that improves AI visibility, not just direct conversion.

Lever 3 — Profile consistency (NAP)

NAP stands for Name, Address, Phone. AI models check whether data is consistent between Google Business Profile, website, hotelleriesuisse database, Booking.com, and TripAdvisor. Inconsistencies are an AI signal for "less trustworthy".

Check specifically:

Is the address identical everywhere ("Strasse" vs "Str."!)
Is the phone number the same on all platforms
Are reception opening hours equally maintained everywhere
Is the website URL exact (with or without www, with or without https)

It's a 2-hour task and one of the most effective invisible levers for AI visibility.

Lever 4 — Google Posts as mini content snippets

Google Posts are small updates directly in the Business Profile. They drop out of the main area after 7 days, but AI models continue to capture them — and use them as a visibility signal.

For hotels that means: 1 post per week on specific topics. Examples:

"Fresh snow in Wengen this week, ski pass and hotel as a package"
"Seasonal tip: Wellness Sunday on 15 June with extended opening hours"
"Direct booking bonus: 10% off the half-board upgrade when booking through our website, until 30 June"

The AI sees: hotel is active, has seasonal offers, communicates directly. That improves the likelihood of being recommended in an AI Overview as a "hotel with current offers".

What no longer works

Much that was SEO standard in 2018 is ineffective or harmful in 2026:

Keyword stuffing in the website description. Building in "Hotel Zurich" 8 times doesn't help with the AI, it gets flagged as spam.
Artificial backlink building. SEO agencies building link profiles do more harm than good.
Generic "top 10 tips" blog posts. The AI prefers specific hotel content, not general travel advice.
Auto-generated content without curation. AI recognises bad AI text and penalises it.

What to do specifically now

This week: check NAP data on 5 platforms and make it consistent (Lever 3)
Next 2 weeks: put 15 FAQ answers on the website with FAQPage schema markup (Lever 1)
From now on, permanent: 1 Google Post per week (Lever 4)
From now on, permanent: collect 4 to 6 new reviews per month (Lever 2)

That's the order of impact. Point 1 is immediately measurable in AI perception. Points 2 and 3 need 8 to 12 weeks until the AI has adjusted. Point 4 is a continuous mechanism.

Where Trophy fits in

Trophy automates exactly these four levers: AI-assisted review replies, automated Google Posts in DE/FR/IT/EN, photo maintenance rhythm, FAQ schema templates for the website. In multi-location setup centralised plus location-specific.

Sources

Innsight Hospitality Industry Report 2026 (AI Overviews local search shares)
Altos Agency 2026 (Direct booking drop analysis)
Visit Mundus 2026 (Hotel AI Overviews Threat Analysis)

AI Overviews are quietly erasing your bookings — what Swiss hoteliers must do in 2026