Can Independent Travel Brands Beat the OTAs in AI Search? What 5,500 AI Answers Tell Us.
For independent travel brands, competing with the likes of Booking.com and Expedia on Google has never been particularly straightforward. Bigger budgets, stronger domains and years of SEO investment have made some of the most valuable search results difficult to break into.
But AI search appears to be playing by slightly different rules.
When someone asks ChatGPT to recommend a safari company or Google’s AI Mode to suggest a specialist tour operator, the answer doesn’t necessarily reflect who’s sitting at the top of Google’s traditional search results.
And despite all the attention on digital PR and third-party mentions as the key to AI visibility, our research suggests that travel brands’ own websites still have a substantial role to play.
We wanted to understand what was actually happening, rather than rely on assumptions about how AI search works.
So we analysed 5,500 AI answers to see which travel brands were being recommended, where the information was coming from and what it might mean for independent operators.
The findings were encouraging, although perhaps not for all the reasons you’d expect.
What we looked at
Between 14 and 22 September 2026, Minty tested 100 UK travel prompts across five AI search tools: ChatGPT, Google AI Mode, Google AI Overviews, Gemini and Microsoft Copilot.
That gave us 5,500 answers, containing 30,711 citations across 3,123 websites. We also tracked mentions of 44 travel brands and compared AI recommendations against traditional Google results for 55 searches.
A few things are worth establishing before getting into the findings.
This is a snapshot of AI search during September 2026, not a set of permanent ranking rules. Perplexity wasn’t included, and our selection of brands and prompts doesn’t represent the entire travel market.
Crucially, we’re looking at visibility, not sales. Being recommended by an AI tool doesn’t necessarily mean someone will click through, enquire or make a booking.
Still, the results tell us quite a lot about how travel brands are being discovered.
1. You don’t have to rank on Google’s first page to be recommended by AI
For years, getting onto page one of Google has been one of the clearest measures of search success.
It’s still important, of course. But it might no longer be the only shortlist worth worrying about.
70.2% of brands recommended in AI answers weren’t among Google’s top ten results for the same search.
That figure comes from the 55 searches where we could compare AI recommendations directly with Google’s first page.
It’s a sizeable difference, and one that should interest any travel brand competing in a particularly crowded search market.
Take a specialist operator offering tailor-made trips to Japan. Ranking for a broad term such as ‘Japan holidays’ might mean competing against companies with considerably more resources.
But when someone asks AI for recommendations based on their interests, preferred style of travel or specific requirements, the resulting selection of brands can look quite different.
That doesn’t mean AI is automatically better at recommending specialists, or that traditional rankings no longer influence the answers.
It means there’s another place where potential customers might encounter your brand, even if you haven’t cracked Google’s top ten.
And right now, plenty of travel businesses aren’t measuring whether they appear there at all.
2. Independent operators are getting a bigger share of the conversation
The next finding makes that opportunity a little more tangible.
When we compared independent and specialist operators against major agents and booking platforms, the difference between Google and AI was striking.
Across our 55 matched searches, independents accounted for 57.6% of tracked agent and operator appearances on Google’s first page.
In AI answers, that figure rose to 83.5% of tracked brand mentions.
That’s a substantial shift in visibility towards independent and specialist businesses within the brands we studied.
And across all 100 prompts, independent and specialist operators accounted for 91.4% of mentions among the tracked agents and operators.
Some of the strongest performers were established specialist names. Abercrombie & Kent appeared 2,473 times, Scott Dunn 1,383 times and Black Tomato 1,347 times.
Booking.com appeared 162 times.
Now, before we start writing the obituary for the OTAs, there’s another number worth considering.
When Booking.com did feature, its average recommendation position was 1.4, the highest of any brand we tracked.
So while it appeared less frequently in our dataset, it was usually very prominent when recommended.
Our prompts and tracked brand selection also matter here. These figures aren’t evidence that independent operators are taking market share from the OTAs or generating more bookings.
But they do suggest that the biggest booking platforms aren’t automatically dominating every AI recommendation.
For travel brands with a clear specialism, that’s an interesting opportunity.
3. The biggest surprise? AI is getting most of its information from brands themselves
If you’ve spent any time reading about AI search recently, you’ve probably come across the argument that third-party mentions are becoming increasingly important.
Get featured in respected publications. Appear in relevant recommendations. Build your brand’s presence beyond your own website.
There’s good reasoning behind that advice, and our findings certainly support the value of external coverage.
But they also tell a slightly different story.
57.3% of the 30,711 citations we analysed came from travel brands’ own websites.
Press and editorial sources accounted for 25.3%, with other sources making up the remaining 17.4%.
In other words, the biggest source of information in our study wasn’t travel journalism or comparison articles. It was the information companies had published about themselves.
And that becomes even more interesting when you look at what people were asking.
For broader travel discovery questions, external sources contributed 45.9% of citations, compared with 48.5% from brand websites.
Pretty evenly matched.
But for more specific questions about which company to book with, the balance changed considerably.
Brands’ own websites accounted for 65.9% of citations.
That makes sense when you think about the information needed to recommend a particular operator. What trips do they offer? Which destinations do they specialise in? Who are their holidays designed for?
Those are questions a brand’s own website should be well placed to answer.
Of course, a citation doesn’t automatically translate into a recommendation. But the volume of references to owned content is difficult to ignore.
So, where does that leave SEO and digital PR?
Rather than treating AI search as a reason to shift attention away from your website and towards third-party coverage, the findings suggest the two need to work together.
A useful destination guide, detailed itinerary or well-explained specialist offering gives AI something substantial to reference.
Relevant editorial coverage can help establish your brand’s presence in the wider conversation around that destination or type of holiday.
Neither should exist in isolation.
For marketing teams, this might mean thinking less about separate SEO and PR targets and more about the specific destinations, experiences and customer questions you want your brand to be associated with.
The same commercial priorities should inform both.
4. Thousands of websites are being cited. But a handful are doing most of the work.
Across the study, AI tools referenced 3,123 different websites.
That sounds like a fairly broad spread of sources.
But when we looked at where the citations were concentrated, a much smaller group stood out.
Just 88 websites accounted for half of all citations.
That’s 2.8% of the websites referenced in our dataset.
The most-cited domains included a mixture of travel brands, specialist publishers and established media.
Scott Dunn topped the list with 950 citations, followed by Which? with 931. Other frequently referenced sites included Condé Nast Traveller, Black Tomato and African Safari Mag.
It’s worth remembering that these are totals across the searches we tested, not a definitive list of publications every travel brand should be targeting.
A specialist cycling holiday company, for example, is likely to care about a different set of sources from a luxury safari operator.
The useful question is which websites consistently appear when AI answers questions about your particular market.
That could make for a much more focused approach to content and digital PR than simply chasing coverage on the biggest publications.
And yes, listicles really do feature heavily
One format appeared particularly often in the results.
Of the citations where we could classify the type of page, 48.1% came from listicles.
These are the familiar articles bringing together recommendations for a particular destination, experience or type of travel.
They’re useful to readers because they put several options in one place. And, based on our data, AI tools reference them frequently too.
We wanted to understand whether appearing in these articles was associated with greater brand visibility.
So we divided our 44 tracked brands into three groups based on how many AI-cited listicles featured them.
The typical brand in the lowest group appeared in three listicles and received 71 AI mentions.
For the middle group, those figures rose to 14 listicles and 179 mentions.
And in the highest group, the typical brand featured in 35 listicles and received 444 mentions.
That’s more than six times as many AI mentions between the highest and lowest groups.
It’s a compelling pattern, but there’s an important caveat.
We can’t conclude that getting featured in more listicles causes AI to recommend your brand more often.
Brands appearing in lots of relevant articles may already be more established or have stronger visibility for other reasons. And when AI references an article that includes a particular company, that company may naturally appear in the resulting answer.
What we can say is that presence in frequently cited listicles is associated with greater visibility in our dataset.
For travel marketers, that makes it worth investigating which articles AI is already using, whether your competitors feature and where your brand might be a genuinely relevant addition.
It doesn’t mean buying your way onto every ‘best of’ list is suddenly a sensible marketing strategy.
AI search is still evolving, and tactics designed purely to influence recommendations may not remain effective. A shortcut that works today isn’t necessarily something to build your long-term visibility around.
5. Your AI visibility depends on where you look
One final finding worth paying attention to is how differently the AI tools behaved.
We often talk about ‘AI search’ as though it’s one thing, but the platforms don’t all draw on the same balance of sources.
ChatGPT, for example, sourced 49.5% of its citations from earned media and other third-party coverage, compared with 44.7% from brands’ own websites.
Gemini looked quite different, with 26.5% of citations coming from earned sources and 65.5% from owned websites.
Google AI Overviews showed a similar preference for owned content, at 66%.
These differences don’t mean one platform is inherently better for a particular brand, or that you should divide your marketing budget according to the percentages.
But they do show why checking your visibility in ChatGPT alone isn’t enough.
You might appear regularly in one tool and rarely in another. The competitors being recommended, and the information supporting those recommendations, may also vary.
Without checking across platforms, it’s difficult to know where the actual gaps are.
What should travel marketing teams do with this information?
There’s plenty of discussion about how to ‘optimise for AI’, often accompanied by a long list of things brands supposedly need to start doing immediately.
Our findings suggest a more useful approach is to start by understanding what’s already happening in your market.
Start with the questions you want to win.
Think about the searches that matter commercially, not just the ones with the highest search volumes. Which destinations, itineraries or specialist experiences bring in your most valuable customers? What do people ask when they’re comparing operators?
See who’s getting recommended.
Test those questions across ChatGPT, Gemini, Google’s AI tools and Copilot. Look at whether your brand appears, how it’s described and which competitors are being suggested instead.
Follow the citations.
Look beyond the names in the answers and examine the websites supporting them. Are there relevant publications featuring your competitors? Are your own pages providing enough useful information? Are there obvious gaps or inaccuracies?
Work out what actually needs fixing.
You might discover that your website doesn’t explain your specialist offering particularly well. Or that a handful of relevant publications are repeatedly cited without mentioning your brand.
Some issues will be straightforward to address. Others will need a more considered content or PR strategy.
Either way, you’ll have a much better starting point than simply deciding you need ‘more AI visibility’.
And once you’ve made changes, keep checking. Mentions and citations are useful indicators, but relevant visits, enquiries and bookings are what ultimately matter to the business.
So, can independent travel brands beat the OTAs?
Our research suggests there’s more room for independent operators in AI recommendations than traditional Google rankings might lead you to expect.
That doesn’t mean the OTAs are losing their grip on travel bookings. We didn’t measure that.
Nor does it mean there’s a single tactic that guarantees your brand a place in an AI answer.
What the findings do suggest is that AI search offers another opportunity for specialist operators to be discovered, and that both the information on their own websites and the coverage they receive elsewhere have a part to play.
For travel marketers, that’s probably the most useful conclusion.
You don’t need to abandon everything you know about search marketing. But you do need to understand how your brand is being represented in a new place where potential customers are looking for recommendations.
Because ranking on Google and being recommended by AI are no longer quite the same thing.
Read the full State of AI Search in Travel report
This article covers some of the headline findings, but there’s plenty more in the research.
Our State of AI Search in Travel 2026 report takes a closer look at the brands appearing most frequently, the websites influencing AI answers, the differences between platforms and what travel marketers can learn from the results.
Where does your travel brand stand?
Knowing there’s an opportunity is one thing. Knowing whether your brand is making the most of it is another.
Minty’s AI Search Kickstarter gives you a clearer picture of where you’re appearing, which competitors are being recommended instead and what you can do to improve your visibility.
We’ll look at the questions that matter to your business, assess your presence across AI tools, investigate the sources behind the answers and put together a prioritised action plan.
