Your Hotel Is Not Behind on AI. It Is Watching the Wrong Clock.

Every hotelier I speak with lately carries a quiet worry that they are already behind on AI, and almost every one of them is afraid of the wrong thing.

I understand the worry, because I have stood on both sides of it. I spent years moving from reservations to the front desk to the manager on duty chair at properties flying flags like IHG and Sheraton, learning the exhaustion of a building that never closes and the small daily heroics nobody on the outside ever sees. These days, I spend my time thinking about where AI belongs inside a hotel rather than how it gets sold to one. From that vantage point, the present moment is not the emergency the market wants you to feel, and it is also nothing. It is something more interesting than either, and understanding it clearly is worth more than any tool you could buy this quarter.

So let me start by defending the skeptics, because they are mostly right.

Your Skepticism Is Not a Weakness

If you are looking at the flood of AI products landing in your inbox and thinking that what you already have works perfectly well, you are not a laggard; you are reading your own operation correctly. A property with a property management system that holds together, a revenue process that fills rooms, and a front desk that knows how to rescue a bad arrival does not have an operational gun to its head. A great deal of what is being thrown at that operational layer is solving problems your hotel either does not have or already handles well enough, and a fair amount of it is last year's mediocre software with a chat box bolted onto the front of it. Being unimpressed by most of it is not resistance to change. It is judgment, and judgment is the scarcest thing in this entire market right now.

The overwhelming feeling so many hoteliers feel is a rational response to noise, not a sign that they are slow. The problem is not that they are failing to keep up. The problem is that the thing they are using to measure their safety is the wrong instrument.

There Are Two Clocks, and You Keep Checking the Wrong One

Here is the part that gets missed, and it is the reason this moment feels so disorienting. Two clocks are running inside your business, and they are not telling the same time.

The first clock is operational. On that clock, there are no forced moments, no deadline, no cliff. Your systems can keep working great for a long while, and the comfort you feel watching them do so is real and earned. Nobody should rip out a functioning stack because a conference panel told them to feel anxious.

The second clock is the demand side. It measures how a traveler discovers, evaluates, and chooses a hotel in the first place, and that clock is moving in a way that has nothing to do with how smoothly your building runs. The way a guest finds you is starting to drift away from a human scrolling through an online travel agency toward an intermediary doing the choosing on the guest's behalf. Soon, a guest will not pick your hotel. Their AI will. A property can run a flawless operation and still quietly lose the moment of discovery if the new layer sitting between the traveler and the hotel cannot read it, understand it, or transact with it.

Quick tip: Once a quarter, ask a simple question that has nothing to do with your operation. If a capable AI assistant were planning a trip to your market and choosing on the traveler's behalf, would it find you, understand what makes you different, and be able to act on it? If the honest answer is no, that gap is invisible on every report you currently read.

The Trap Is That Nothing Breaks

This is what makes the moment so easy to misjudge. The threat does not announce itself, because nothing in the building breaks. You will not see it on a daily flash report or a morning stand-up. It shows up as a slow erosion of direct visibility that everyone blames on the OTAs, or on a soft market, or on the weather, when what has really changed is the route demand takes to reach you.

Your operations clock says everything is fine, and your operations clock is telling the truth, and that is precisely why it is the wrong thing to be staring at. Comfort on the operational side is not the same as safety on the demand side. The two simply are not connected, and treating one as proof of the other is the most expensive quiet mistake a hotel can make right now.

The Real Cost of Your Silos

This is also where the silos and the legacy systems matter, but not in the way the vendors pitch them to you. The disconnected stack and the data trapped inside it are not an operational emergency. Your hotel ran fine on them yesterday and will run fine on them tomorrow. What they cost you is the ability to respond quickly once the demand shift becomes undeniable, because by then the integration work is a prerequisite rather than a calm project you chose to start on a quiet Tuesday.

The real cost of the silos is optionality, and optionality is the thing you only notice you lost at the exact moment you needed it. You pay that bill later, under pressure, at the worst possible time, which is how most hotels end up making rushed and expensive decisions that serve a vendor far better than they serve the guest.

Quick tip: Treat your data and your integrations the way you treat the roof. You do not replace the roof because it is exciting. You maintain it so that nothing catastrophic happens the day a storm finally arrives. Clean, reachable data and open connections are the roof work. Unglamorous, easy to defer, and brutal to retrofit in a downpour.

Two Decisions, and Most Hotels Make Them in the Wrong Order

If you asked me where the money should go, I would tell you to stop treating investing in AI as one decision, because it is two, and most hotels are making them backwards.

There is a foundation layer, and there is an application layer. The application layer is the chatbots and the copilots and the upselling engines, the shiny things in the demos and the slick booths at the trade shows. That layer is volatile and getting cheaper every quarter, which means almost anything you commit to heavily today will be outclassed or simply absorbed into something else within a year or two. On the application layer, the honest advice is to rent rather than marry. Experiment cheaply, stay unattached, and let other people fund the expensive lessons.

The foundation layer is the opposite in every way. It is your data being clean and reachable instead of being buried in an export nobody on your team can use. It is your system being able to talk to the outside world through open interfaces. It is your people being fluent and unafraid when they work alongside these tools. None of that looks like an AI purchase, none of it photographs well in a demo, and all of it holds its value no matter which application vendor eventually wins the war you are being asked to bet on today. That is the investment that serves a hotel for years. It is also ignored, precisely because it is slow and quiet and impossible to put on a banner.

Quick tip: Before any AI spend leaves the account, make it answer one question. Does this measurably improve the guest experience or free up human attention for the moments that matter? Anything justified only by fear of falling behind, or by a promise of cutting headcount, should be killed on the spot. Fear and cost-cutting are the two motivations that reliably produce the wrong investment at the wrong time.

The Sentence Everyone Repeats at Conferences

There is a line you have heard a hundred times by now, usually delivered as hard-won wisdom. It says you will not lose your job or your market share to AI, but to the people and brands who learn to use AI better than you do.

I believed it myself for a while, so let me take it apart gently, because it sounds like the sober take while smuggling in the same old urgency, wearing a better suit. The true half is genuinely true. The threat is human rather than technological, and judgment about the tool matters far more than mere access to it. But notice the quiet assumption underneath it, that the contest is really about AI proficiency. That puts AI back at the center of the story and turns buying the tool into mastering the tool, while keeping the tool itself as the prize.

A hotel does not win because its staff are better at prompting than the hotel down the street. It wins because it never lost the plot about what it was optimizing for, and used whatever tool served that. So, I would keep half of that saying worth keeping and quietly retire the rest. The threat is human, yes. But the human skill that decides the outcome is not AI fluency. It is clear what is worth protecting in the guest experience and being willing to use these tools to protect it, while your competitors are busy collecting AI skills for their own sake. The brands that win this will not be the best AI users. They will be the ones who treated AI as a means and never once confused it for the point, which was always the part of hospitality that made anyone want to pay us in the first place.

Most of What You Are Being Sold Is Surface AI

Here is the distinction that changed everything for me, and the one I wish more of this industry would sit with. Most of what vendors describe, and to be fair, what a lot of hoteliers describe right back, is what I have come to call surface AI. It is the layer you can see without any effort, the demo and the dashboard, and the confident promise on the slide, and it runs almost entirely on a carrot and a stick. The carrot is the efficiency, the revenue, and the time you will supposedly get back. The stick is the fear of being left behind. Between the two, surface AI rarely teaches anyone anything at all. It simply confirms whatever they already believed. The optimist leaves more excited, the skeptic leaves more skeptical, and the perceptions quietly fulfill themselves while everyone congratulates each other on the demo.

I say this with some humility, because I did not understand AI either until I stopped staring at the surface and started diving below it. That is not a figure of speech. It is below the surface, in the actual mechanics of how these systems reason and fail and occasionally surprise you, that a different gateway opens, the one that leads to your own reality of what is really happening rather than the version someone built to sell you something. You cannot reach it from a demo. You only get there by going under, and once you do, you stop being a spectator to the technology and start being able to judge it.

And here is what was waiting for me down there, the part that matters most to anyone who has ever run a hotel. What hospitality really means is determined by feelings, not algorithms. No model measures the relief of an exhausted guest who finds a room ready early, or the quiet loyalty earned by a manager who fixes the thing before anyone has to mention it. The surface will always try to convince you that those moments are data points to be optimized. They are not. They are the whole reason the building exists, and the only sane purpose for any tool you ever bring near them.

This Is Why I Work Where AI Meets Blue Ocean

None of this makes me pessimistic about technology, and I want to be clear about that, because it is where my whole perspective comes from. Used below the surface and used well, AI and Blue Ocean thinking do the same quiet thing. They enhance five things at once: your capability, your creativity, your capacity, your innovation, and your productivity. They give a small team the reach of a much larger one and give a tired operation the room to think clearly again. But that enhancement is never the prize on its own.

The prize is what you choose to do with the room it opens, and for me, the answer never changes. You build and amplify the skills that AI cannot replace: judgment, warmth, the instinct for a guest that no model will ever hold. The tool expands what your people are capable of. Your people remain the entire reason any of it is worth doing.

Going Below the Surface Means Closing the Skill Gap

Getting below the surface is not as technical or as intimidating as it sounds, and from where I sit, it is the single most important thing a hotelier can do right now, to educate themselves and learn to understand AI in the right context. AI today is not a contest between people and machines. It is a collaboration between them, and the value lies in how well the two are combined rather than in technology on their own.

This is also where the danger hides, because a skill gap left open does not stay neutral. It quietly turns into the very thing people fear. If your team never learns to work alongside these tools, the perception that AI is a threat hardens into reality, not because technology came for their jobs but because nobody helped them learn to use it. The fear becomes self-fulfilling, and the hotel ends up living out the story it was most afraid of.

So let me be precise about what I believe will happen. I do not think AI will replace jobs. I think it will substitute certain functions and tasks within a job, which is a very different claim, and the entire difference between the two lies in whether your people have the skill to absorb the change. A front desk agent does not disappear. The twenty minutes of copy and paste in the middle of their shift might. What they do with the time that opens up is the whole game.

The shape of that skill is easier to teach than most people assume, and I think of it as the 4Ds.

Delegation is deciding what work belongs to humans, what belongs to AI, and how to distribute the tasks between them. It draws on the same instinct a good manager uses, building a shift rotation, knowing your goals, knowing what each side is capable of, and making a deliberate choice about who covers what rather than letting it happen by accident.

Description is communicating clearly with the system. It means defining the output you want, guiding the process, and being specific about the behavior and the tone you expect, which is a skill any hotelier already owns from briefing a new hire on exactly how a guest should be greeted.

Discernment is the critical part, evaluating what the AI gives you rather than accepting it. It is assessing quality, accuracy, and whether the answer is even appropriate, and deciding where it falls short. It is the same muscle a revenue manager uses when reading a competitor report and knowing instantly which numbers to trust and which to question.

Diligence is using these tools responsibly and ethically. It means making thoughtful choices about which systems you let near your guests, staying transparent about how they are used, and taking full accountability for AI-assisted work as if you had done every word of it yourself, because in the eyes of the guest, you did.

Close that gap, and the fear dissolves on its own, because a team that knows how to delegate, describe, discern, and stay diligent is not threatened by the tool. They are the reason it works.

What I Would Actually Do This Year

If I were sitting across from you with a coffee, I would not hand you a shopping list. I would tell you that you are not behind, that you do not have to choose a tool this year, but that you do have to decide this year whether your hotel will be ready to choose well later, and ready to be found at all once the choosing moves out of human hands.

That readiness is rarely a solo project, and it is rarely a pure technology project, which is why every hotel taking this seriously needs an AI champion somewhere in its orbit. Not a vendor selling you a platform, and not a consultant who has never worked on a night audit, but someone who has stood at the front desk and also understands what these systems can and cannot do, someone whose first loyalty is to the guest and the team rather than to the tool. That is the work I care about most and the partnership I most enjoy, helping hotels build the foundation and the human readiness that lets them move with confidence instead of panic when the moment finally forces a decision.

If that is the conversation you want to have, I would rather have it before the storm than during it. The tools are the part you are allowed to wait on. The readiness is not.

Frequently Asked Questions

Is AI a must for hotels right now?

Not operationally. A hotel with working systems is not facing an emergency, and the skepticism many owners feel about the flood of AI products is healthy. The shift that matters is on the demand side, in how guests discover and choose a hotel, and that is the part worth preparing for early because it is hard to work under pressure.

Will AI replace hotel staff?

I do not believe AI will replace jobs. I believe it will substitute certain functions and tasks within a job, which is a very different thing. A front desk role does not vanish, but some of the repetitive work inside it might, and the real question is whether your team has the skill to use the free time well. The parts of hospitality that earn loyalty, judgment, and warmth in the moment something goes wrong, do not delegate cleanly to a machine, and any AI investment justified mainly by headcount reduction is usually the wrong investment for the wrong reason.

What should a hotel invest in first when it comes to AI?

The foundation before the applications. Clean and reachable data, open integrations, and staff who are fluent and unafraid hold their value no matter which tool wins. The flashy guest-facing products are the volatile layer that gets cheaper and better every quarter, so rent there rather than marry.

How will AI change the way guests book hotels?

Discovery is starting to move from a human scrolling an OTA toward an AI intermediary choosing on the traveler's behalf. A hotel can run perfectly and still lose visibility if it is not legible and reachable to those systems, and that loss shows up slowly rather than as anything that obviously breaks.

What is an AI champion, and why would a hotel need one?

An AI champion is someone who understands both the floor of a hotel and the capabilities of these tools, and who keeps the guest at the center of every decision rather than the technology. The role exists to help a property build readiness and make calm, well-timed choices instead of fear-driven ones.