How to Choose an AI Consultant Who Understands Hotel Operations, Not Just Technology
Why does an AI consultant need to understand hotel operations?
Because hotel AI projects rarely fail on the technology; they fail on the operation wrapped around it.
Take a guest asking your new website chatbot whether they can bring their dog. The chatbot pulls from the website FAQ, which says yes in two room types. Your Booking.com listing says no pets. The rate notes in the PMS mention a £25 charge, and the reservations team will tell the guest it depends which room is free. That's four answers to one question, and the guest only needs one of them to be wrong to arrive with a spaniel you can't accommodate.
An advisor who understands hotels spots that the problem is four versions of the truth and fixes that first. One who doesn't sells you a better chatbot, which now gives the wrong answer faster and more politely.
The chatbot was never the problem. The four different dog policies were.
What's the difference between an AI advisor and an AI vendor?
A vendor is paid when you buy their product. An advisor is paid for judgement, and the clearest test of that judgement is whether they'll ever tell you not to buy something.
The line gets blurred because plenty of people sit in between: resellers, "implementation partners" and consultants earning commission on the tools they recommend. That isn't automatically wrong, but you should know which one you're talking to before the recommendation arrives rather than after you've signed. Ask directly: "Do you receive commission, referral fees or partner incentives from any tool you might recommend to us?" A straight answer takes five seconds. Anything longer is an answer too.
What questions should a hotel ask an AI consultant before hiring them?
Generic checklists tell you to ask for case studies and references, and you should. These questions go further, because the answers show whether someone has worked inside a hotel or only sold to one.
"If I told you we lose enquiries after 9pm, what would you want to know before suggesting anything?" A good advisor answers with questions: who covers nights, where calls divert, who checks the voicemail and when, how many of those calls become bookings. A weak one starts describing their voice agent.
"What would you tell a hotel not to do?" A good advisor has a specific answer, such as not automating guest replies until your policies match across every channel. If they can't name anything, everything is for sale.
"Tell me about a time two systems held different versions of the truth. How did you handle it?" Listen for specifics, like rates in the PMS not matching the channel manager or a spa diary nobody else could see. "It plugs into everything" isn't an answer.
"Whose job changed because of your work, and how?" You want a named role and both sides of the change: what that person stopped doing and what new checking they picked up. "It freed the team to focus on guests" with nothing behind it is a brochure line.
"What happened the last time something you built went wrong in front of a guest?" Hospitality has no back office for mistakes; the guest sees them. A good advisor has a real story and can tell you what they changed afterwards. "It hasn't happened" means they haven't built much, or aren't telling you.
"Who owns this after you've gone?" There should be a named internal owner, a written process and training that survives the next round of staff turnover. If the honest answer is "us, on a retainer", you're renting rather than building.
"Which operational roles have you actually held?" Not a dealbreaker on its own, but someone who has run a front office knows things about handovers and check-in peaks that no amount of tool knowledge replaces.
What questions should a good AI consultant ask you?
The other half of the test is what they ask you. In a first conversation, an advisor who understands hotels should be asking questions like these before they recommend anything, and if they don't, they're preparing a pitch rather than a diagnosis.
What happens to a phone enquiry that comes in at 9pm on a Friday, and who picks it up on Monday?
Which systems hold your rates, availability and policies, and which one is right when they disagree?
Where does the same information get typed in twice, and who does the typing?
Which reports does the GM need every morning, and how long do they take to put together?
How would this change the front desk agent's shift, or the night auditor's?
What's your staff turnover like, and who would own a new process once the person who learned it has left?
What would a guest notice if this went wrong?
A consultant who asks about your 9pm phone calls is diagnosing. One who demos a voice agent is selling.
What are the red flags that an AI consultant is selling technology rather than solving operational problems?
The first meeting is a demo. Diagnosis should come before prescription, every time.
They promise savings before they've seen your data. "Save 20 hours a week" from someone who hasn't looked at your rota or your systems is a guess dressed up as a number.
Every problem has the same answer. If missed calls, slow reporting and inconsistent rates all lead back to one product, the product is the point.
They talk about "the hotel" and never about a person. Operations are run by people in specific roles, and advice that never mentions one hasn't been near an operation.
Your head housekeeper wouldn't follow the pitch. If the people whose day changes can't understand what's being proposed, adoption will stall and the budget will quietly disappear.
The knowledge stays with them. No documentation, no training and no handover plan means you're buying a dependency.
Should a hotel hire a generalist AI consultant or a hospitality specialist?
For most hotels, a specialist. Generalists know the tools, but hospitality has a set of conditions that change how every one of those tools behaves: a 24-hour operation, shift handovers, seasonal teams, guests who notice everything and a tech stack usually assembled from four or five vendors that were never designed to talk to each other.
A hospitality specialist knows that a "simple" automated morning report touches night audit, revenue and the GM's 9am meeting, and plans for all three. A generalist finds that out three weeks into the project.
A useful test: if the advice you're being given could apply equally well to a marketing agency or a software start-up, it isn't hotel advice.
What should the first engagement with an AI consultant look like?
It should start with readiness rather than tools. Before anything is bought or built, a good advisor works out where you actually stand: how enquiries flow and where they leak, which systems hold which version of the truth, how much manual reporting your team carries each week and how comfortable they are with change.
The output should be a prioritised plan you could hand to anyone, including a vendor of your own choosing. The first piece of work should be small and tied to a commercial outcome you can measure, like enquiries answered within the hour, missed calls converted or the three hours reservations spend compiling the Monday report.
The uncomfortable part is that an operator-led advisor will sometimes tell you things you'd rather not hear. Your biggest AI problem is often a data problem. Some workflows will change and some roles will look different in a year. Hearing that early is cheaper than discovering it after the contract is signed, and it's also the clearest sign you're getting advice rather than a sales process.
A good advisor will cost you a difficult conversation. A bad one will cost you a year.
Why I built FAI Consultancy this way
I spent nearly 20 years in hospitality before AI became my day job: General Manager roles, multi-site sales and marketing, and stints running housekeeping, the restaurant, front office and commercial across the UK, Europe, Asia and Australia. I still work inside an operating hospitality business, so the 9pm enquiry problem isn't theoretical to me. I also sit on the AI Hospitality Alliance's AI Readiness Framework committee.
That's why every FAI engagement starts with the operation. The AI Readiness Assessment maps where your business actually stands and what to fix first, implementation is tailored to your systems rather than a standard stack, and training gives your team the confidence to run it after I've gone.
If you recognised your hotel somewhere in this article, it's fixable. Book a 15-minute intro call, or if you'd rather start on your own, the free resources at are a good place to begin.
Key takeaways
Choose an AI advisor who asks about your operation before showing you any technology.
Ask every advisor directly whether they earn commission or referral fees on the tools they recommend.
Judge a consultant as much by the questions they ask you as by the answers they give.
A good advisor can tell you what not to do and who will own the work once they've left.
Start with an AI readiness assessment and a small first project tied to a measurable commercial outcome.
If the advice could apply to any business, it isn't hotel advice.
Frequently asked questions
How do I know if an AI consultant understands hotel operations? Describe a real problem to them, such as an out-of-hours phone enquiry or a rate disagreement between your PMS and channel manager. Someone with operational understanding asks about your roles, systems and handovers before suggesting a tool.
Should a hotel AI advisor be independent of technology vendors? Ideally, yes, or at least fully transparent. Ask whether they earn commission, referral fees or partner incentives on anything they recommend, and get the answer before the recommendation.
What should a hotel do before hiring an AI consultant? Get clear on the operational problems costing you money or time, such as missed enquiries, slow response times or manual reporting. Then look for an advisor who starts with an assessment of your readiness rather than a product pitch.
Is a hospitality-specialist AI consultant better than a generalist? For most hotels, yes. Hospitality runs around the clock on fragmented systems and shift-based teams, and a specialist plans for those conditions from day one rather than discovering them mid-project.
