Guests have stopped searching for hotels

I have been working on AI visibility for properties in a few different categories lately, and the same pattern keeps appearing.

It is easiest to see in wellness, where the product is complicated and the guest rarely knows the right words for what they want.

Someone types "wellness hotel Greece" into Google. They get a page of links.

They open five or six sites, each of which greets them with a headline about philosophy, a gallery of the spa, and a list of treatments with names they do not recognise.

Twenty minutes later they are no closer to a decision than when they started, and the properties have begun to blur into one another

The reason this fails is that almost nobody thinks in product categories. A person planning this kind of trip is not holding the phrase "wellness retreat" in their head. They are holding something closer to this:

• I have not slept properly in eight months, will a week actually fix that or am I fooling myself?
• I want to go with my husband, but he hates anything organised. Is there somewhere we can both be comfortable?
• Why is one place three thousand euros and another twelve thousand for what looks like the same thing?
• I had surgery in spring. Is it safe for me to do this, and will anyone there actually check?
• Do I have to join group activities, or can I simply be left alone?
• Who is in charge of my programme, a doctor or a receptionist with a timetable?

Until recently, the only places to get those answers were a phone call with the reservations team, or scattered articles that stayed on the surface. Most people did neither. They guessed, booked, and hoped.

A guest no longer works through ten websites. Much of what they used to establish during a first enquiry call, they now settle before it.

They open an assistant and describe their situation in ordinary language: "I am completely burnt out, I have five days in October, I want somewhere by the sea in Europe, quiet, with proper food. Where should I go?"

And they receive two or three names.

For a property, the question becomes whether it can be one of those names. That is not a matter of visibility in the old sense.

For an assistant to recommend you with any confidence, it has to be able to connect six things:

1. the guest's actual problem,
2. the way your programme addresses it,
3. your property specifically,
4. the expertise behind that programme,
5. the people delivering it,
6. and some evidence that any of this is true.

If one link in that chain is missing from everything written about you, the answer names somebody else.

All of them now have to work with intent rather than with a service name.

Not "our spa offers these twelve treatments", but "I am considering doing something serious about my sleep and I am trying to work out what that would involve and whether it would suit me".

1. The practical work starts in an unglamorous place. I collect the questions guests genuinely ask, usually a hundred or more, from reservations calls, enquiry emails, reviews and the spa team's memory.
2. Then I sort them. Some are noise. Some are asked once and never again. Perhaps ten to fourteen of them are the questions that actually decide whether someone books, and those become the structure that everything else is built around.

Classic SEO answers the question of where to find a wellness hotel in Greece. It produces a list, and the guest does the rest of the work themselves.

An assistant is answering something else entirely: what should I do about my situation. To appear in that answer, being visible is not enough. What is written about you has to be the thing the person was hoping to hear.

The side effect is the part that interests hoteliers most, in my experience. This work does not only improve how you appear in AI answers. It improves the website for everyone who lands on it.

The people arriving are a closer fit, they stay longer, and the enquiries that arrive are better, because someone writes to you after they have understood that the place suits them, rather than while they are still comparing a dozen tabs.

So the honest description of the job is not writing nicer copy. It is assembling an accurate picture of what people actually arrive with, and then rebuilding the site, the articles, the social content and the practitioners' own material around that picture.

The sensible first step is usually a diagnostic: where the property currently stands in these answers, what is missing, and what is worth doing first.

My approach is deliberately focused: identify the gaps that matter, improve the highest-value areas first, and measure whether further investment is justified.

See how my GEO optimization service works →

If any of this sounds like your property, let’s talk.

On the call I will explain how this work is put together and what it involves, show you with examples what changes on the site and beyond it, and answer your questions.

Then I will propose something for your particular situation: what is worth doing, in what order, and what a sensible starting point looks like.

A little about me. I have been in marketing for over twenty years and I now work with independent, wellness and countryside hotels, mainly on direct bookings and visibility in AI search.

Among recent results: confirmed paid bookings arriving from ChatGPT and Perplexity for a five-star property, with no advertising budget behind them at all.

I also built my own tool for measuring whether AI assistants name a property or name its competitors, across several dozen of the questions its guests genuinely ask.