Your hotel's strategy most likely involves OTA commission, a direct booking target, a RevPAR figure, a paid search budget and a goal for ancillary spend per guest. It almost certainly does not include data on whether AI assistants recommend your property.
That is an understandable gap, because until recently the data was not there to measure. It matters more than it first looks, though. When an AI assistant cannot find a clear answer on your own website, it describes your hotel from wherever else it can find one: OTA listings, old reviews, directory entries and third-party mentions you may not even know exist. That second-hand picture is what guests act on, and it feeds decisions well beyond marketing. Here are three places it shows up.
.png)
How your property shows up in AI recommendations tells you a lot: how it is positioned, what the market thinks it offers, and where the gaps are. Most hotels have never been able to see this before.
Your hotel's strategy most likely involves OTA commission, a direct booking target, a RevPAR figure, a paid search budget and a goal for ancillary spend per guest. It almost certainly does not include data on whether AI assistants recommend your property.
That is an understandable gap, because until recently the data was not there to measure. It matters more than it first looks, though. When an AI assistant cannot find a clear answer on your own website, it describes your hotel from wherever else it can find one: OTA listings, old reviews, directory entries and third-party mentions you may not even know exist. That second-hand picture is what guests act on, and it feeds decisions well beyond marketing. Here are three places it shows up.
- Take a capital investment. You approve a spa extension and expect it to lift wellness bookings and treatment revenue. The business case assumes guests will learn the spa exists. But if AI assistants still describe your hotel as a comfortable business property with a small gym, because that is what they find online, fewer guests discover the spa and the payback takes longer than planned. From the inside, that is hard to explain.
- Take channel strategy. Shifting bookings away from the OTAs depends on guests reaching you directly, before anyone else shapes their choice. AI assistants now sit at that early stage. Unlike the OTAs, they are not covered by any agreement, and they appear in almost none of your reporting.
- Take positioning. Assistants describe hotels in words as well as ranking them. If several of them keep calling your resort family-friendly while you are investing to reposition it as an adults-only luxury retreat, the market is hearing the wrong story, and that is worth catching early.
We have opened a free pilot of the PQ Impact Hotel AI Visibility Report to hotels in any destination, so you can base these decisions on measurement instead of guesswork. This article covers the patterns we see most often, what the report includes, and what taking part involves.
1. The one area with no reporting line yet
Almost every part of hotel performance shows up in a report someone reviews each month. This one does not, at least not yet.
When a guest follows an AI recommendation to a competitor, they never reach your website. There is no bounce rate, no abandoned booking, no drop in session time. Your dashboards look normal, because you lost that guest before they ever touched the systems you watch.
Revenue teams react to changes in the numbers, and here there is nothing to react to. A hotel can lose part of its discovery traffic over a whole season with no obvious signal. That is why the problem tends to surface late, usually as soft direct bookings that nobody can quite explain.
The pilot report fills that gap. It shows how often four assistants recommend your hotel against your direct competitors, measured rather than guessed.
2. AI recommends the hotels it can describe
Assistants have never used your spa or eaten in your restaurant. They work only from what is written about you online, and they favour information that is specific and consistent across sources.
This has a real consequence. A modest hotel with clear, well-structured, consistently named content can be recommended ahead of a better hotel that describes itself in vague, atmospheric language. "World-class wellness facilities" gives a model almost nothing to work with. "2,600 sqm spa with eleven treatment rooms, an indoor thermal circuit and a Turkish hammam" gives it plenty.
We see this in most audits. The hotels winning recommendations are often the most clearly described, even when they are not the strongest in the market.
3. The outlets that drive ancillary revenue are often the least visible
Rooms usually show up reliably. The facilities that carry your margin often do not.
Some examples we see again and again: a professional tennis academy reduced to one line in a Booking.com amenity list; a Relais & Châteaux membership buried in a website footer where no crawler counts it; a restaurant listed under three different names across the website, the OTAs and Google Business Profile, which splits it into three weak listings instead of one strong one.
In each case a real strength never reaches the guest asking about it. Ask an assistant which resort on the coast has serious sports facilities rather than a games room, and the hotel with the full academy can lose to one with two courts and clearer, better-structured content.
The report shows which of your strengths assistants currently struggle to read, grouped by type of guest question, so you can see exactly where what you offer and what gets understood come apart.
4. A baseline makes the next decision easier
There is a lot of GEO and AEO advice around now. Some of it is well founded, some of it is inferred rather than tested. Most of it is generic, which makes it hard to apply to your specific hotel with any confidence.
Acting on generic advice without a baseline means you can spend on the wrong problem. We have seen hotels commission entire content programmes when the real issue was their name being inconsistent across four platforms, roughly a fortnight of work to fix.
A clear starting measurement makes it far easier to decide what deserves budget and what does not.
5. There is a window while the field is still open
In most of the destinations we index, very few hotels have started on this. So acting now, rather than in eighteen months, is likely to make a real difference, and it will cost less to do today than to catch up later.
The early movers in Google search built positions that took rivals years to challenge. The same pattern looks like it is forming here, on a shorter timeline.
What the free pilot gives you
The full report, built with the same engine and the same depth as our paid work. Free while the pilot runs, in any region.
What we measure
We never ask a model whether your hotel is good by name, because branded questions just produce flattering answers that tell you nothing useful. Instead we ask what real travellers ask, using neutral prompts built for the destination and its categories. Then we record which hotel gets named, where it ranks and how it is described. Ten prompts, thirty snapshots, four assistants: ChatGPT, Claude, Gemini and Perplexity.
What you receive
- Share of voice across four AI assistants. How often you are recommended versus your direct competitors, one assistant at a time.
- Performance by guest type. Where you stand across the question categories that drive revenue: luxury, honeymoon, family, wellness, business and location-led searches.
- Competitor benchmarking and sentiment. Which hotels are winning the recommendations you want, why they are winning them, and the exact language assistants use to describe your brand.
- A prioritised twelve-point action plan. Ranked fixes across content, structured Q&A, schema markup and entity clarity, written plainly enough for your own team or your agency to action directly.
You get it as a secure report link within 72 hours, once your destination is indexed. If your market is not indexed yet, we build it and confirm timing when you sign up.
What we ask in return
Fifteen minutes. A short call after you have read the report, telling us which findings were useful and which were not, plus permission to include your destination in our index. No contract, no payment details, and nothing about your hotel is ever published without your approval.
What happens afterwards
Entirely up to you. Keep the report and act on it in-house. Ask us for a competitor report if that helps. Move to ongoing monitoring or hands-on implementation only if you want to. Or do nothing more, with no chasing from us.
Questions hoteliers usually raise
- Do travellers actually book through AI assistants? Most do not, and that is a fair point. Bookings still happen on a website, an OTA or a metasearch site. What has changed is the discovery stage, where guests form their first shortlist. If you are not on that shortlist, the later stages never happen.
- Our guest profile is older. Does this still apply?Usually yes, just indirectly. Adult children and travel advisors often do the research for them. And AI answers now appear inside normal Google results, so even guests who have never opened an assistant still see them.
- We already invest in SEO. How is this different?SEO is about how travellers find you through a search engine. This is about how assistants answer open travel questions, and whether you appear in the answer. The technical groundwork overlaps in places, but how you measure visibility, and how you lose it, are different.
- How can we judge whether the findings are reliable? Check them against what you already know. That is part of why the pilot is free. Read the findings next to your own knowledge of the hotel and the market, then decide whether the work is worth taking further.
Find out where your hotel stands - Sign up to pilot
If your destination is already indexed, your report follows within 72 hours. If it is not, we confirm timing when you sign up.
