AI search is
the next OTA battle.

Ask any hotelier where the money leaks, and the answer comes back fast: commission. Every booking that arrives through an online travel agency carries a cut, and that cut is the single largest recurring cost most properties never quite manage to shrink. For ten years the industry has fought the same fight, loyalty schemes, book-direct campaigns, rate-parity battles, all aimed at one goal: winning the guest back to the hotel's own website, where the booking is worth far more.

That fight is not over. It has simply moved to a new arena, one where the rules are still being written and where very few hotels have shown up yet.

The commission every hotelier already resents

The economics are simple and brutal. A booking that comes through an OTA can cost the property somewhere between fifteen and twenty-five percent in commission. The identical booking, made through the hotel's own site, costs only a few percent in payment processing. On a room that sells for a few hundred dollars a night, that difference is real money returned to the business, every night, on every stay.

This is why direct bookings are the most valuable revenue a hotel has, and why an entire decade of industry effort has gone into shifting share from the agencies back to the property's own channels. It remains the single biggest margin lever in the business.

The front door is moving to AI

Here is what has changed. When a traveller decides where to stay, the first move is increasingly not a search box full of blue links. It is a question put to an AI system: best boutique hotel on the Gold Coast, somewhere to stay near Byron for a weekend, a quiet place close to the conference centre. The answer that comes back is not a page of ten options to compare. It is a short, confident recommendation, sometimes a single name.

Google's AI answers now sit above the traditional results in Australian search. Answer engines return a property directly, or they don't. And when they name a hotel, they either send the guest toward the hotel's own site or toward an agency listing. That choice, made by the machine, decides who captures the margin.

Cited means booked direct

This is the whole game in one line. If the AI can read a property's own site, understand it, and trust it enough to recommend it, the guest lands on the hotel's booking page and the property keeps the margin. If the AI can't, it falls back on what it can read and trust, which is almost always the agency listing, and the commission clock starts again.

The agencies are already optimising for this

The uncomfortable part is that the other side is not standing still. The large travel agencies have the resources, the engineering teams, and the commercial motivation to make their listings the source AI systems reach for. Every month a hotel waits, the machine's habits harden a little more around the agency's pages rather than the property's own.

That is the cost of doing nothing. Not a sudden drop, but a slow default, where the AI's picture of "where to stay in this town" quietly settles on whoever was easiest to read and trust first. Reversing an established preference is far harder than earning the position while it is still open.

What it takes to be the source AI trusts

Winning the citation is not mysterious, but it is different work from traditional marketing. AI systems don't rank a list; they build a picture of a business and decide whether it is clear, consistent, and trustworthy enough to name. That picture depends on whether the property's own pages can be read cleanly by a machine, whether the essential facts, name, location, rooms, offers, are stated plainly rather than buried in imagery, and whether the same details line up everywhere the business appears online.

None of that is exotic. But it is rarely in place by accident, and it is almost never audited. A property can look polished to a human visitor and still be close to invisible to the systems now shaping the first recommendation a guest ever sees.

The unsettling part is how often this is invisible to the operator. The site looks fine, the booking engine works, the photos are beautiful, and the crawler that decides the recommendation quietly never gets a clean read. Nobody tells you. You simply stop being the answer.

Why the timing is the whole point

The properties that establish themselves as the trusted answer during this window, while AI habits are still forming, are likely to hold that position for a meaningful stretch afterward. This is a first-mover advantage with a closing door, and the door is closing at the exact moment buyers are changing how they choose.

For a single property it is a margin decision. For a group operating dozens of hotels, it is the same decision multiplied, one strategy, many properties, and an ROI story that writes itself the moment even a small share of bookings shifts from agency to direct.

Most tools measure whether you are visible. The more useful question, and the one worth answering before your competitors do, is whether you are the one getting cited by the AI at the moment the guest decides.