Will the Machine Mention You?

Somewhere right now, a hotelier is doing everything right.

Her property is beautiful—genuinely, expensively beautiful. The reviews hover near five stars. The spa is serene. The restaurant is excellent. The photography is polished. The rates are competitive. Twenty years of hospitality wisdom says she has won.

Last night, three hundred miles away, a woman planning her honeymoon opened ChatGPT and typed: "We want somewhere on the Riviera Maya that doesn't feel like everywhere else. Somewhere with a soul. Where should we stay?"

The machine thought for two seconds and recommended a handful of hotels. The hotelier’s wasn't one of them and she has no idea.

And that's the point.

The decision happened without her.

The Gate Moved

Every generation of hospitality has had its gatekeeper. First it was the travel agent, armed with a Rolodex and opinions. Then came Google, with ten blue links and a knife fight for page one. Then the OTAs aggregated everything, ranked everyone, and charged fifteen to thirty percent for the privilege.

Each shift moved the decision a little further away from the hotel itself. But there was always a consolation: you could still be found. Maybe a traveler discovered you through Condé Nast Traveler. Maybe through a Google search. Maybe through an OTA list, a friend's recommendation, or an article written years earlier. The path wasn't always direct, but it existed. Given enough curiosity, a traveler could eventually find their way to you.

That consolation is disappearing.

Forty-three percent of travelers already use AI to plan trips, and the channel is growing fifty percent faster than traditional search. AI doesn't have a page four. It doesn't ask travelers to compare twenty tabs. It doesn't return hundreds of results.

It answers with a few names and a few reasons. And increasingly, that's enough.

You are either in the answer or you do not exist.

The implications are larger than most hoteliers realize. A recent study of more than 131,000 properties across thirty countries found that only about sixteen percent of hotels are visible to ChatGPT, Google's AI, and Perplexity at all. Of those that are visible, seven major hotel groups dominate the top twenty-five brands.

Think about what that means. The largest brands have always had the largest budgets. What they didn't always have was control of the conversation. Now they do.

Meanwhile, many independent hotels—the very properties built on individuality, character, and story—are becoming invisible. Not because they are worse. Because they are harder to describe.

The Machine's Secret

Most people assume the machine recommends the best hotels. It would be comforting if that were true. But the evidence suggests otherwise.

Researchers recently ran more than 4,500 travel prompts through ChatGPT across nine destinations and discovered something unexpected: the relationship between review scores and AI visibility was surprisingly weak. In other words, excellence alone is not what gets you recommended. Even celebrated, awarded properties routinely go missing from the answer.

When you think about it, this makes perfect sense. The machine isn't choosing hotels. It's choosing explanations.

Someone asks a recommendation. The AI doesn't simply produce a few names, it has to justify them too. It needs a sentence. A reason. Something it can say.

A property built around a clear idea hands the machine its answer: a wellness sanctuary rooted in the sacred feminine, a converted monastery overlooking the sea, a design hotel inspired by local mythology. The machine immediately knows what to say.

But what about the property that is unquestionably beautiful?

Beautiful architecture… Beautiful rooms… Great service… Fantastic beach… Delicious restaurant… Gorgeous pool…

The problem is that describes half the luxury market.

The machine does not recommend the best hotels. It recommends the hotels it can describe.

And those are not always the same thing.

What a Sentence Is Worth

Historically, every new discovery channel took a larger share of the transaction. Travel agents took their commission. OTAs took fifteen to thirty percent. Intermediaries inserted themselves between guest and property and collected rent for doing so.

What's interesting about AI is that it may be the first major discovery channel whose default behavior points people directly to the hotel's own website. Not Booking.com. Not Expedia. Your website.

That matters.

Last year, OTAs accounted for sixty-one percent of independent hotel bookings. Cancellation rates on those bookings approached fifty percent compared to roughly eighteen percent for direct bookings. Direct guests also generate significantly greater lifetime value, often five to eight times the value of their first stay.

The economics are difficult to ignore.

There has always been a distinction between demand you own and demand you rent. Demand that arrives by name is demand you own. Demand that arrives through comparison is demand you rent.

The machine didn't create that reality. It simply made it more obvious.

The Photograph the Machine Can't See

Now comes the strangest part of the story.

Pictures.

For all the anxiety surrounding AI, today's recommendation engines are surprisingly blind. They read far better than they see.

When the machine assembled that honeymoon shortlist, it wasn't looking at photographs. It was reading language: reviews, articles, captions, blog posts, comments, Reddit threads.

And nobody writes long posts about another infinity pool at sunset.

We've all seen too many of them.

People write about ideas. About symbols. About experiences. The image that stops someone mid-scroll. The ritual they can't stop talking about. The detail that becomes a story. SAbout things they haven't seen before.

Those stories become articles. Those articles become language. And language becomes training data.

The machine cannot see your photographs. It can only read what your photographs made people write.

An image that generates no language is, to the machine, an image that does not exist.

And the pressure is coming from another direction as well. AI has made beautiful imagery effectively free. Anyone can generate it. Everyone is generating it. Endless streams of it are converging into the same visual language.

At the same time, Google's Gemini already processes images alongside text, and visual search is among its fastest-growing frontiers. Soon machines will not only read. They will look. And when they do, a hundred nearly identical luxury photographs may register as variations of the same image.

For more than a century, beauty was hospitality's most valuable currency. Now it is being manufactured at scale.

Beauty is free. Meaning is the moat.

The Property With a Sentence

I watched this happen years ago, before AI entered the conversation.

A new resort arrived on the Riviera Maya carrying two difficult labels in luxury hospitality: vegan and all-inclusive. Beautiful property, but beautiful properties are everywhere. The coastline is full of them.

What mattered was not the architecture. It was the idea:

A wellness sanctuary built around the sacred feminine.

Not a tagline. Not a marketing campaign. An organizing principle. Something that shaped the imagery, the language, the rituals, the guest experience, and ultimately the perception of the property itself.

Then the idea did what strong ideas do.

People started asking for the hotel by name. Branded searches increased tenfold within seven months. Paid social campaigns generated returns approaching six hundred percent. The internal team carried the foundation forward to earn two Michelin Keys and recognition from Condé Nast Traveler and much more.

That wasn't an SEO victory. It wasn't a media victory.

It was a positioning victory. People could explain why the property mattered.

The machines are simply enforcing the same rule. If a person—or an AI—can say why you, you get chosen. If they can't, you get compared.

Tonight

Try something.

Open ChatGPT. Ask about your hotel. Then ask for the best properties in your market. Read the answers carefully.

Not the rankings, the language.

If the description of your property could just as easily describe your competitor, you've found the problem. And it probably isn't your website. It probably isn't your SEO. It probably isn't your photography.

Those things matter. But they aren't the root issue.

The root issue is simpler.

What is the one compelling and unique sentence a machine would repeat about you? Something people have begun to repeat on their own, through all the channels.

Because increasingly, that sentence is becoming the difference between being chosen and being invisible.

The hotels that make up the visible sixteen percent are not necessarily better. They are more describable.

And that is a choice.

Conclusion

In the end, the machine is forcing hospitality to confront a question it has been avoiding for years: why this place?

Not what it has. Not what it looks like. Not what it costs…

Why does it exist?

Answer that provocatively and clearly enough and the answer spreads through everything: the architecture, the imagery, the rituals, the guest experience, the conversations people have about it, and ultimately the phrasing the machine repeats.

The story comes first. Everything else is its shadow. 

That’s why I’m here. Learn More