Skip to content
Travel

How travellers pick a hotel inside an answer

A traveller resolves one trip question into six comparisons, and the two that settle it, the cancellation terms and what is included, are the two hotels publish least readably.

5 min read · Published 2026-10-15

A hotel is shortlisted on what the engine could read about it, and most of that is written by a booking platform. By the time a traveller sees a rate, an assistant has already resolved their question into six comparisons: the location and what is nearby, the room and what is included, the price for their dates, the cancellation and refund terms, what recent reviews say, and whether booking direct is cheaper than a platform. A property whose rooms, inclusions and policies sit only inside a booking engine gets described from a platform listing, or left out of the shortlist entirely.

Six questions, asked in one breath

A prompt like best hotels in India for a family holiday with kids does not stay one question. It fans out into six: which area suits the trip, what the room includes and whether it sleeps the family, what a night costs on these dates, what cancelling costs and by when, what guests have said recently, and whether the direct rate beats the platform. The assistant answers all six from whatever it can read, then writes one shortlist.

The rate is the last thing it reaches, not the first. A traveller who opens a booking platform has usually already been given a shortlist by an answer, and the platform is where a price gets attached to a decision that was mostly made.

Almost none of the material the assistant reads has to come from the property. Booking platforms, review sites, travel publishers, maps and forums all describe a hotel whether or not its own site does, and the platform's version is the one most often cited.

Why do the cancellation terms decide the shortlist?

Because a rate is a headline and the cancellation window is the actual promise. A property whose policy lives inside the booking engine, three screens past the room selection, is a property an assistant has to guess at. It does not skip the question. It takes a platform's restatement of the policy, which may be a season old, or it infers one, and sometimes it infers in the property's favour, which is the worse outcome: the guest cancels on the day they believed was free and is charged a night.

Inclusions have the same failure in a different place. A room page says breakfast is included, the platform says it is not, and the confirmation email says it is for two guests. Nobody wrote a false sentence. The engine built one out of three true ones, and it stated the most generous.

Airlines carry a version of this in fare rules. A fare that is changeable for a fee on the airline's page and non-refundable on the platform's is two fares to an engine, and the answer quotes whichever it read.

A rate without its dates is not a price

A starting-from price with no season, no date and no occupancy cannot be compared with anything. The traveller asked about specific dates, the platform quotes those dates, and a careful engine prefers the number it can qualify.

That is the mechanism behind the complaint most hotels arrive with, which is that the answer quotes the platform's rate and not theirs, and sends the guest to the platform to book. The platform was not preferred. The platform was legible about dates.

The opening is the same shape as the problem. Publish what a night costs for which dates, or a live path that takes dates and guests and returns a real price, and state the direct-booking benefit outright. A sentence that says booking direct includes breakfast and a later checkout is a sentence an engine can quote in answer to is it cheaper to book a hotel direct or through a booking site.

What does an assistant do with a location it cannot place?

It places the property using somebody else's data. Maps, area guides and itineraries tell the engine what is nearby and how far the airport is, and a property page that says centrally located and nothing else adds nothing to that.

Travellers ask about location in concrete terms: walking distance to the beach, minutes from the station, whether there is a pool the children can use. A page that answers those in words, with distances, is a page the engine can quote for the area question, which is usually the first of the six to be settled.

Reviews are the decide step, and they are the one source a property cannot write. What a property can do is make the facts a review would confirm easy to find, so the answer reads the review and the page as agreeing with each other rather than as a claim and a correction.

The six places the answer finds its evidence

An answer about a hotel, an airline or a platform is assembled from sources, and only one of them is yours. Hotel, airline and platform pages supply rooms, cabins, inclusions and policies, when they are readable. Booking platforms are where rates, availability and restated policies are filtered, ranked and quoted. Review sites carry the recent guest and passenger reviews engines weigh at the decide step. Travel publishers and guides supply the area and timing context. Maps and location data supply what is nearby. Forums and social posts are where a trip that went wrong, or right, becomes public evidence.

You control one of those six directly. The rest you influence by being accurate and easy to verify, and by keeping the facts a platform restates about you identical to the facts on your own page, so there is one version to read.

So a platform listing that states last year's cancellation window is not a listing problem. It is a sentence that will be repeated in answers given to people who never open the platform.

How do you know what the answer says about your property?

You read the answer. There is no proxy for it, and the booking figures will not tell you, because the guest who was shortlisted elsewhere never arrived, and the one who booked through a platform arrived with a commission attached.

So measure it as text. Run the prompts travellers actually type, such as your brand cancellation and refund policy or best travel booking sites for cheap flights and hotels in India, across ChatGPT, Gemini, Perplexity and AI Overviews, and capture the answer, your position in it, and the domains cited behind it. A wrong deadline or an invented fee stops being a suspicion at that point and becomes a sentence you can act on.

The progression to aim for is a ladder. The answer compares places to stay and you are absent. You are named among options, with no detail attached. Your own page is cited as the source for a policy or an inclusion. Your property is in the side-by-side the traveller is deciding from. Your property is recommended as the fit for this trip.

Each rung is earned by publishing what the rung below it was missing: the location and access as text, each room with what it includes in words rather than icons, rates with their dates or a live path to them, the policies in plain words with deadlines and fees, and a booking path that takes dates and guests and returns a real price and a confirmation.

In Ansyra

Measure this for your own hotel, airline or platform

Ansyra runs the prompts travellers actually type across every AI engine on your property or service, records which sources each answer was built from, and shows where rivals and platforms are named instead of you.

Straight answers

Frequently asked

Our reviews are better than the hotel the answer recommends. Why are we not named?
Usually because the engine could qualify their facts and not yours. Reviews settle the decide step, but a property only reaches that step if the answer could first place it, price it for the dates and read its cancellation terms. A property whose policies and inclusions live inside the booking engine is dropped before the reviews are weighed. Publish those as text and the reviews start to count.
Should we publish our cancellation policy in full? Will that not cost us bookings?
It costs you fewer bookings than the alternative. If the policy is not readable, the assistant takes a platform's restatement, which may be stale, or infers one, and a guest who believed a date was free and was charged a night writes the review that costs the next ten bookings. A published policy also lets the engine describe your property precisely, which is the condition for being placed in a side-by-side.
Our rates change by date and season. How can that be text?
Publish the structure, not every rate. State what a night costs in each season with the dates that season covers, say outright what booking direct includes, and keep the booking engine for the exact figure. An engine can quote a seasonal rate and point to the direct path; it cannot read a calendar widget, so a page with only a widget is a page with no price on it.
What is the smallest change that moves this?
Put the cancellation and refund terms, with their deadlines and fees, and each room's inclusions on a readable page outside the booking engine, and keep them identical to what the platforms state. That single change removes the guessing from most answers about your property and makes your own site the cited source for a policy, which is the rung where a property stops being merely named and starts being quoted.

See where you stand in AI answers

Run your real questions across ChatGPT, Gemini and Google's AI answers, free for 14 days. No connections required to start; all five AI engines on every plan.