AI visibility is table stakes for travel brands now, and if you’re not, here’s where to start.
But let’s say you are. Someone finds your product through one of those engines and goes looking for more. What happens next?
What we see is that many brands are not built at that time. Media planning, targeting, moderation: everything was designed for the consumer that started with Google. Here’s what you really need to build an AI-powered one:
Travel has more entry points than your media plan accounts
Travel search is no longer a single channel. A person planning a trip doesn’t start with Google, and scroll down the list of results. They ask an AI chatbot for an itinerary, watch travel content on YouTube, check Reddit for some reliable hotels, and then come back to search when they’re ready to book.
The problem is that most campaigns are not designed for a journey that includes these many touch points. They work independently, without visibility into what other channels are doing, so no single campaign knows enough about a particular user to respond appropriately.
What a demand capture layer needs to look like
So what does it really look like to build campaigns that can see a full journey? It starts with understanding that not all AI-targeted users are in the same place when they arrive.
They may still be researching, going to other channels to confirm what the AI told them, or going straight to a named search query to find it directly.
Each method requires a different response than your typical acquisition campaign, and each setup is different.
For users who are still in research mode, the key channels are where validation happens: programmatic media across the open web, paid community on forums where travel content resides, and Reddit where there is a volume of meaningful discussion around your product. The goal at this stage is not conversion. Presence and involvement. Demonstrating consistently enough that a good first impression from an AI transfer is not eroded by a ubiquitous competitor in your absence. And taking that next step with the user by providing them with valuable information as they look to confirm what they’ve already learned.
For users going to search, the opportunity is how you tag and segment them. If someone clicks from an AI engine to your site, that referral source is traceable. You can use it to build audience segments, and feed that signal back into your paid search campaigns so that those users get different bids, different landing pages, and different messages than a cold visitor. They have already been told that your product is worth considering. The campaign should know that.
The piece that brings the two together ensures that those users are seen as the same person across all touch points. Someone who came in via an AI referral, saw a YouTube ad pop-up, and then searched for your product directly is one user at the end of a short journey. Without building the right audience locally, your campaigns treat them as three unrelated strangers and bid accordingly.
Why moderation stops when parties are silenced
This is where the structural problem becomes more expensive: if you don’t measure this as a connected system, you will never see where it breaks.
A typical setup for most tour types is paid search, paid social, and each program reports its own platform numbers. Google says it has moved the X reservation. Meta says it called Y. The DSP of the program has its own calculation. Add them and you will get double or triple the number of your original booking, because all three platforms require the same user credit.
That’s not a new problem, but it becomes a bigger one when AI adoption happens at the top of it all. None of those platforms can see the transfer of ChatGPT or AI Overview. So a user who has been recommended your brand by an AI engine, searched for you, and booked is said to be using fully paid search. The role of all other channels in that journey is not apparent.
Editing does not change attribute models. Create a measurement framework designed to see the full path: a definition of the development platform on the fly, a marketing mix model for a long-term view of what really drives revenue, and an escalation test to distinguish what really drives bookings from what just wants credit.
The ascending piece is more important in the movement than most stages because the basic purpose is so high. People who were going to book a holiday were I will book a holiday. The real question that paid media needs to answer is not whether the campaign reached the person who booked it. That the campaign created bookings with your brand that would not have happened otherwise. Running a geo-based test, where you use paid media in other markets and hold them with comparables, often produces results that are more sobering than brands expect. A reasonable share of what gets paid media attention turns out to be an ever-changing demand. And most importantly, the right answer to that is not to get the money back. It moves it towards touchpoints that actually create a new need rather than a pre-existing harvesting intent.
The system must be built together
Investing in AI visibility is the right move. But transferring that need to a paid media program that isn’t built to meet it, is the equivalent of running a brilliant product campaign and sending people to a broken booking engine. Awareness did its job, but everything else on the ground did not.
Holding a growing share of travel in the next few years means building the whole system together: The visibility of AI into a paid media structure designed for the needs of multiple points of entry, measured in a framework that can really distinguish what really drives the growing bookings.
That requires strategy, channel implementation, and scale to be owned in one place. It is not linked to many agencies. It is not aligned to the quarterly review. Designed together from the ground up, with a shared view of where users are coming from and what drives them to book.
