A quick note on focus before you read. AuraScope works with businesses preparing for AI-mediated discovery across categories. As the platform matures we are going deep on travel, where AI search, recommendations and agentic booking are reshaping the buyer journey faster than anywhere else. The lessons below apply broadly. Our product focus is travel.
This is not a Google story. It is the discovery layer of the internet reorganising itself, and Google just made it impossible to ignore.
For years travel brands competed for clicks. Then for rankings. Now the contest is inclusion: whether ChatGPT, Gemini and Claude name you when someone asks them where to stay, what to do, how to get there. Those three are the new gatekeepers. They sit between your brand and the traveller, and they decide who gets recommended.
That is the shift AuraScope was built for, and Google's latest Search announcements confirm the operating model we have been preparing travel brands for since the start of the year.
Here is the uncomfortable part. If an AI system cannot read you, verify you and cite you with confidence, it recommends someone else. There is no penalty notice, no ranking drop you can diagnose in an SEO tool. You simply stop appearing, and you may never know why.
Google's Search I/O 2026 announcements make the direction explicit. AI Mode has passed one billion monthly users, the search box is being rebuilt for longer and more conversational questions, and people can carry their context straight from an AI Overview into a full AI Mode conversation (blog.google).
This matters for travel because nobody plans a trip with one keyword. They ask the messy, loaded questions they would ask a friend who knows the place:
"Where should I stay near Queenstown for a family trip, quiet room, parking, late check-in, easy access to day trips?"
That is not a search query. It is a decision brief. And the brand that answers it best, across every source the model trusts, wins the booking.
From search results to decision systems
Google's own engineering guidance explains the mechanics, and it is worth understanding rather than guessing at. Generative features in Search stay grounded in Google's index, but they run on retrieval-augmented generation and query fan-out. One traveller question quietly becomes many related searches behind the scenes (Google for Developers).
So the work changes shape. You are no longer optimising one page for one phrase. You are making sure the facts about your brand are visible, consistent and credible across the whole source graph a model pulls from to build an answer.
For travel that graph is wide: your own site and booking engine, Google Business Profile, the OTAs, review platforms, destination guides, blogs, local press, images, video, schema, and every review a guest has ever left. The model reconciles all of it.
When those sources disagree, the model hesitates. When your direct booking path is unclear, it routes the traveller somewhere else. And when an OTA describes your property better than you do, the OTA becomes the more useful source. You have handed your own story to a reseller.
AEO is not a hack layer
There is a wave of "AEO hacks" coming. Most of it is noise, and Google says so plainly: no special AI markup file, no need to shred your content into tiny chunks, no rewrite of everything for the machines (Google for Developers).
We agree with the spirit of that, and we draw a hard line on what AEO actually is. It is not a trick to game a model. It is the discipline of making a brand legible to AI search before disintermediation lands: visible, citable, and genuinely actionable inside an AI-mediated decision.
For travel that means answering the way a real traveller, or the agent acting for them, would actually ask. Not thin content. Not generic destination filler. Not "luxury accommodation in New Zealand" stamped across ten near-identical pages. It means specific, verifiable, lived-in detail:
- Who is this place genuinely best for, and who should book elsewhere?
- What makes the location actually useful, not just close to things?
- Which room fits which traveller?
- What do guests consistently praise, and what do they complain about?
- What nearby experiences change whether it is worth staying here at all?
- Can a traveller book directly, with confidence, right now?
Generic copy is invisible to a model reconciling dozens of sources. Specifics are what get cited.
The next visitor to your site is not human
The biggest part of this shift is agentic search, and it is the part most brands are not ready for.
Google is rolling out Search agents that monitor information, compare options and act on the user's behalf, and it is extending agentic booking across local services and experiences (blog.google). The agent will not just answer "where should I stay?" It will shortlist, check constraints, weigh trust signals, look at availability, and push the traveller toward a booking.
We build on a simple principle: treat AI agents as first-class citizens. An agent is now as important a visitor to your website as a human, so you design and engineer for it with the same seriousness, not as an afterthought.
That is not abstract. Web.dev spells out how agents read a site: screenshots, raw HTML, and the accessibility tree (web.dev). If your booking flow leans on unstable layouts, hidden interactions, vague buttons or unlabelled form fields, the agent cannot work out what to do, and it moves on. An inaccessible booking flow used to be a compliance problem. Now it is lost revenue.
Why travel feels this first
Travel is the whole journey, not just a hotel room. Flights, the stay, the experiences, the vibe of a place, the kind of person you become when you travel. It is high-intent, high-consideration and full of messy human constraints, which is exactly the kind of decision AI is being built to carry.
Travellers do not only compare price. They weigh location, reviews, room fit, cancellation terms, dietary needs, parking, transfers, family suitability, accessibility, weather and timing, all at once. Google's commerce announcements are heading straight at this. Universal Cart and the Universal Commerce Protocol are being positioned as the rails for agentic purchasing, and Google says UCP is moving into hotel booking, with Australia in the upcoming checkout expansion (blog.google).
So visibility on its own stops being enough. The bar moves to visible, citable and bookable. Can an agent confidently recommend you and send the traveller straight to your own booking path, instead of a reseller's?
The brands that win will be the ones that can see why they are included, why they are left out, which sources are shaping the answer, and what to change. Waiting for traffic to fall is not a strategy. Hoping your OTA presence carries you is not a strategy. Treating AI search as a reporting dashboard is not a strategy.
That is the work we do. We show travel brands why AI recommends a competitor instead of them, then fix the sources behind that answer.
The future of travel discovery will not belong to the brand with the most pages. It will belong to the one AI can understand, trust, cite, and send the traveller to directly.
Where to start
If you are already a customer, or you are just starting to wonder how AI search is affecting your brand, we can assess where you stand in AI answers today.
Our deepest product work is now focused on travel: hotels, operators, and the businesses that move people around the world. Because in travel, being visible is no longer the finish line. Being chosen is.
References