AI search is rewriting how guests find vacation rentals. Will yours get picked?
- Jul 3
- 7 min read
For 20 years, the game was simple: rank on Google or win the search-and-filter battle inside Airbnb. That game is now changing faster than almost anything the short-term rental industry has ever seen.
More and more guests are skipping the search bar entirely and asking ChatGPT or Gemini, "Where should I stay?" On the Host Planet Podcast, we sat down with Amirali Mohajer, Co-Founder of hostAI, to unpack what that shift really means – and why it could be the biggest direct booking opportunity independent operators have ever had, if they move now.
The shift: from searching to asking
In the 'old' world, a guest ran the searches themselves – one query at a time, tab after tab, filtering and comparing until something matched their intent. In the new world, they talk to a single AI assistant that has already crawled and indexed the web behind the scenes, and draws on what it learned in training to answer.
When it doesn't have enough information, the assistant breaks the request into more specific searches, gathers more, and answers. As Amirali puts it, the discovery and planning that used to happen in the guest's own head – and inside the OTA interface – is moving upstream, into the AI layer. That relocation of effort is the whole opportunity.
What actually makes a listing "legible" to an AI
So when an LLM picks one property over another, why? Amirali breaks it into three layers.
1. Can it read you at all? Before anything else, the assistant has to be able to access and read your listing. Big platforms – Airbnb, Booking, Vrbo – have machine-readability baked into their DNA. Independent operators are behind: your website has to work for two audiences now, the humans who visit it and the AI systems deciding whether to surface you. Making your storefront easy for AI to crawl and understand is, in his words, "table stakes – what you need just to start the conversation."
2. Are you relevant? Next, the assistant narrows to a relevant subset using the guest's profile, memory, and everything shared in that session. His example: a bachelorette party of eight heading to Nashville is a perfect match for one five-bedroom rental and completely wrong for another. This contextual matching used to happen in our own minds as we sorted and filtered. Now we've outsourced it to the AI – a genuinely new dynamic.
3. Can it trust you? Finally, the assistant decides which options it's comfortable putting in front of the guest. Big brands lean on brand power. Independents build reliability by having authoritative third parties talk about them – reputable magazines, tourism boards, local governments, credible writers. In effect, outsiders underwrite your reputation online. This is where PR and citations become a revenue lever, not a vanity exercise.
Amirali is candid that the mechanics will keep evolving as these platforms compete on user trust. But the one thing that won't reverse is the shift of context from humans to AI – and the move toward long-tail discovery happening at the assistant level rather than on the OTA.
Why AI search favours direct bookings
Here's the encouraging part. Because the AI re-ranks results by relevancy and trust, you no longer have to be the number three or four result on Google. You can be result number 10, 15, even 20 in the initial crawl, and still be surfaced after the assistant applies its own layer. HostAI is already seeing the momentum: roughly a 40% month-on-month increase in AI-generated traffic to the direct booking sites it operates.
But traffic is only half the job. "Can you persuade them to book directly with you?" Amirali asks. Many arrivals never knew direct booking was an option and had an OTA in mind. You have to create enough trust, and offer payment options seamless enough, to convert them once they're through the door.
The five-dial combination lock
Amirali describes direct booking as a five-dial combination lock – acquisition, proposition, trust, payment, and retention – where most hosts get two or three dials right and still watch it fail. All five have to align before the lock opens.
Two dials get missed most often. The first is proposition – being genuinely clear about why a guest should book with you rather than the platform. The second is retention, which he considers badly under-invested. Too many operators never collect guest data, or they collect it and do nothing, or they fire off generic marketing blasts at the wrong moment. Given that most people take about two trips a year and spend a few weeks planning each, there are only four to six weeks in the entire year when your message is even relevant – so what you send and when matters enormously.
A related, easily-missed dial is price-advantage data. Operators set a direct rate and an OTA rate and assume the job's done – but then stack promotions on the OTA to win ranking, quietly eroding or erasing the direct discount exactly when high-intent guests are deciding.
The hidden cost of OTA dependence
Everyone knows the commission. Fewer people calculate the real cost: the lost guest relationship. Amirali's challenge is worth trying yourself – call 10 guests who booked through an OTA and ask if they can name the brand they stayed with. In his experience, about 8 in 10 can't. That means you re-acquire the same guest next time and pay commission again and again. Worse, unremembered brands get commoditised, which lowers the value of your business in the eyes of owners you're pitching to manage.
The underestimated shift: AI as your distribution team
Most of the headlines focus on discovery. The part Amirali thinks is under-discussed is using AI agents to level up your own distribution. Building a website, running SEO, running search ads, keeping direct pricing competitive with OTAs – this once meant thousands of pounds and several agencies that never quite co-ordinated. AI can now automate that go-to-market motion at a scale that was previously cost-prohibitive for an independent brand. For the OTAs, AI is an optimisation of something already strong; for independents, it's a genuine chance to close the distribution gap.
Where to start (and the honest order)
If you're starting from zero, the timeline depends on your foundations – how distinctive your properties are, how strong your reviews and brand are, how long you've held a domain and built a presence. Amirali's recommended progression is clear: storefront → retention → organic → paid.
Start with the storefront, and build it for guest acquisition, not owner acquisition – show owners you can drive yield rather than just telling them. Then drive the right traffic to it (helped by the "billboard effect," where guests discover you on an OTA and seek you out directly). Next, treat your past guests as an asset and nurture the relationship without being pushy. Only then layer in organic marketing you own, followed by paid, which buys a fast boost but stops working the moment you stop paying.
And no, you probably shouldn't aim for 100% direct. With rare exceptions – think highly distinctive inventory paired with an owner who already has a large personal audience – Amirali suggests direct should be around 50% of your booking mix, with OTAs kept as an acquisition channel whose guests you then nurture back to direct.
The bottom line
AI search isn't a distant trend; it's already reshaping discovery. If your property only exists inside an OTA, you're becoming invisible to a growing group of travellers who never open Airbnb – and the window to get ahead is open right now. Get your storefront machine-readable, earn third-party trust, close all five dials, and start counting the hidden cost of every guest you don't own.
Listen & explore
Hear the full conversation with Amirali Mohajer on the Host Planet Podcast, brought to you in partnership with Lodgify – the all-in-one software giving independent hosts a direct booking website, channel manager, and guest messaging without needing a tech background. Use code HOSTPLANET20 for 20% off. Click here for details. Catch the full episode on YouTube, Spotify, or Apple.
Want to go deeper on AI-powered direct booking infrastructure? Explore hostAI at gethostai.com.
FAQs
How do ChatGPT and Gemini decide which rental to recommend? According to hostAI's Amirali Mohajer, AI assistants filter on three layers: whether they can access and read your listing at all (machine-readability), how relevant you are to the guest's profile and context, and how much they can trust your brand – trust built through authoritative third-party mentions.
What is the first thing an operator should do to get found by AI search? Make your website machine-readable so AI systems can crawl and understand it. This is considered "table stakes" – big platforms like Airbnb and Booking already have it; independent operators need to catch up.
Does AI search help independent hosts win direct bookings? Yes, potentially. Because AI re-ranks results by relevancy and trust, you don't need to be the top Google result. HostAI reports roughly a 40% month-on-month rise in AI-generated traffic to the direct-booking sites it operates – but you still need to convert that traffic with trust and seamless payment.
What are the five dials of direct booking? Acquisition, proposition, trust, payment, and retention. Most hosts get two or three right and still fail; all five must work together. Proposition and retention are the most commonly missed.
What is the hidden cost of OTA dependence? Beyond commission, it's the lost guest relationship. Around 8 in 10 guests can't recall the brand they booked through an OTA, so operators re-acquire the same guests repeatedly – and unremembered brands become commoditised, lowering business value.
Should a host aim for 100% direct bookings? Usually not. Amirali recommends direct making up around 50% of the mix, keeping OTAs as an acquisition channel whose guests are nurtured back to direct – except in rare cases of highly distinctive inventory with a built-in audience.
Comments