AI is reshaping travel distribution. Holiday parks are building their own.

More and more travellers are starting with an AI assistant instead of a search engine or booking site. For decades, the distribution model has been pretty well understood: travellers search, large platforms aggregate demand, and operators pay commission for access to that demand. AI changes that.

For travel industries, that raises a really important question: do we just accept whatever the next distribution model looks like, or try to shape some of it ourselves?

My family owns and operates two holiday parks in Australia. We know OTAs provide value, and we use them. But we also know the trade-offs all too well. So if travel distribution is going through another major shift, it’s worth asking whether industries like ours have an opportunity to do things a little differently this time.

My wife and brother-in-law, at our park in Launceston

Commission for reach: the deal we accepted

OTAs became dominant for good reasons. They made fragmented accommodation easier to find, compare and book. They also gave operators access to travellers they might never have reached themselves. The bargain was simple: give up a percentage of the booking in exchange for distribution.

Over time though, that relationship has become… complicated 💔 Brand bidding is an obvious (and somewhat damning) example. A traveller searches Google for a business by name. The business may have spent years building their brand and attracting that customer, but an OTA can advertise against the search and potentially turn that demand back into a commissioned booking, effectively hijacking customers and selling them back to the businesses that built the product.

For holiday and caravan parks, there’s another issue too: most large accommodation platforms were fundamentally designed around hotels. A powered site for a 9-metre caravan, a pet-friendly cabin, or an ensuite site is not another hotel room.

The platforms work, and we use them because that is what’s available. But they weren’t built for our industry, and the deal could be better. So why not make a new one?

Many industries have their own 9-metre caravan

Holiday parks make this pretty easy to illustrate. A traveller might ask:

I’ve got a 9-metre caravan and a dog. Where can I stay near Byron Bay?

It sounds simple, but answering it correctly requires quite a bit of information. You need site dimensions, pet rules, availability, minimum stays, access restrictions and seasonal rules. Sometimes you also need to know the traveller’s actual setup - tow vehicle, awning, slide-outs, whether the dog is allowed on that particular site, and so on.

Other travel sectors have their own version of this problem: tours, cruises, activities, regional transport, specialist accommodation. There are always details that matter a lot to the customer that might not fit particularly well into the dominant booking schema. The more specialised the product, the harder it is to squeeze it into someone else’s model without losing useful information.

Discovery no longer has to start in a marketplace

One of the interesting things about AI is that discovery doesn’t necessarily have to start inside a marketplace. A traveller can describe the trip they’re trying to take and an assistant can potentially pull information from several different sources before recommending where to stay or what to book. That means competing for distribution doesn’t necessarily mean building another Booking.com.

For some industries, it may be more useful to become the best source of information about their category - then make that information available to AI assistants, search engines, publishers, industry websites, and other partners. And that requires owning some of the underlying pieces: structured data, availability, rules and booking pathways.

Selling shovels in the AI gold rush

This is one part of the current AI shift that is easy to overlook. When a major technology change happens, the opportunity isn’t always the obvious product sitting directly in front of it.

AI is producing new travel assistants and new booking interfaces almost every day now. But there are also real opportunities underneath the hood. Who has the best data? Who describes the product properly? Who provides live availability? Who becomes the source that all of these new interfaces rely on? This is the particularly interesting bit.

A shared distribution layer, without the commission

With Total Parks, we’re building a shared distribution layer for the Australian and New Zealand holiday park industry. In practice, that means taking data from different park management systems, normalising it, adding the park-specific information travellers actually need, and making live availability and booking available through different channels. Those channels can be Total Parks itself, an industry or tourism website, or an AI assistant.

The aim isn’t to replace every existing distribution channel. We want to give the industry another one - and importantly, one where parks don’t have to give up a percentage of every booking. For us, AI isn’t just another marketing channel. It changes what is technically and commercially possible.

How Total Parks distributes park availability and bookings through different channels, without taking a commission on every booking.

Making industry websites bookable

This is where industry and tourism bodies become interesting. Australian state caravan and holiday park bodies already run consumer websites, maintain park directories and spend money attracting travellers. Those channels already exist. They’re already industry-owned.

So rather than sending that demand somewhere else to transact, why not make the industry website bookable? That’s something we’ve started doing. The first Total Parks integrations are already live with industry bodies in Victoria and Tasmania.

The Total Parks booking engine (in the red box) built into the Caravan & Residential Parks Victoria website.

An individual operator can improve its own website and booking funnel. But an industry working together can potentially own more of the infrastructure sitting underneath discovery and booking.

Holiday parks are our example, but plenty of other travel sectors could make the same argument.

A window before the new model settles

I’m not suggesting the major OTAs are about to disappear. They have enormous advantages: brand, demand, capital, technology and consumer trust. AI may strengthen some of those businesses just as easily as it creates opportunities for new ones.

AI distribution yields mixed results

But the next version of travel distribution isn’t settled yet. Travellers are changing how they search. AI platforms are still working out how travel discovery and booking should work. The interfaces and commercial models underneath all of this are still moving. That gives industries a window to get their own data in order, build useful infrastructure and make sure they’re part of what comes next.

What should your industry own?

For me, this is the more interesting question. As AI changes travel distribution, what should an industry actually own? The data? The customer relationship? Its own distribution infrastructure? The machine-readable rules and information that describe what makes its products different?

Holiday parks are our test case. If the distribution model is changing anyway, industries shouldn’t automatically assume somebody else is going to build the next one for them.

In short: AI is changing how travellers find and book, the next distribution model isn’t settled, and industries that get their data and infrastructure in order now have a chance to shape it. Holiday parks are doing that with Total Parks.

Dan Wild is the Director of Total Parks, based in Launceston, Tasmania. He’s a software engineer who has lived and worked in holiday parks for over a decade. To learn more about Total Parks, visit totalparks.com.au.