There is a growing assumption in travel marketing that AI visibility is a single metric, you either show up or you don't. Our research suggests the reality is far more nuanced, and for some brands, far more uncomfortable. Being mentioned by AI frequently is not the same as being recommended. And the gap between those two things is where a significant amount of commercial opportunity is being lost.
To understand this gap, we ran 200 real purchase-intent queries through Claude, ChatGPT, and Perplexity, the three AI platforms UK consumers are increasingly using to research and book holidays. We tracked which brands appeared in each response, and crucially, which brand each platform named as its single top recommendation. The contrast between two brands in our study tells the story better than any summary can.
The Numbers That Should Concern Every Travel Brand
On the Beach appeared in 159 of our 600 responses, the sixth highest total in the study, ahead of every brand outside our dominant top five. By any traditional measure of AI visibility, that looks like a strong result.
But of those 159 mentions, just 13 became a top pick recommendation. That is an 8% conversion rate from mention to recommendation.
159 On the Beach total mentions across 600 responses
13 Times named as the outright top pick
8% Mention-to-recommendation conversion rate
On ChatGPT, On the Beach was mentioned 42 times. It was chosen as the top recommendation once.
Now consider loveholidays. It appeared in just 57 responses, less than a third of On the Beach's total. Yet of those 57 mentions, 40 became top pick recommendations. That is a 70% conversion rate. On Perplexity specifically, the platform that draws on live web data rather than training data alone, loveholidays converted 23 out of 24 mentions into a first-place recommendation. A 96% conversion rate.
57 loveholidays total mentions
40 Times named as the outright top pick
70% Mention-to-recommendation conversion rate
loveholidays has fewer than half the mentions, and more than three times the recommendations. That is not a marginal difference. That is a fundamentally different relationship with AI.
What AI Actually Says About On the Beach
To understand why the conversion rate is so low, we looked at the language AI uses when it does mention On the Beach. The same phrases appear with striking consistency across all three platforms:
- "Good for comparing deals"
- "Often competitive"
- "Worth checking"
- "Specialises in beach holidays"
- "Competitive prices and flexible payment options"
"On the Beach, good for comparing deals across different operators" is not a recommendation. It is a consolation prize. The brand is being positioned as a tool to check, not a place to book.
This language is what we call utility positioning. AI has learned to describe On the Beach as a useful instrument in the holiday research process (a comparison layer, a price-check resource) rather than a destination with a clear identity and a strong reason to choose it over alternatives. Utility positioning appears in lists. It rarely appears at the top of them.
Compare that to how AI describes loveholidays when budget, value, or popular sun destinations are involved. In those moments, AI doesn't hedge. It doesn't say loveholidays is "worth checking." It says loveholidays is the answer.
The Perplexity Problem Is a Signal, Not a Coincidence
On the Beach's performance on Perplexity is the most revealing data point in the study. On Claude, it appeared 87 times. On Perplexity, that dropped to 30, a 66% collapse.
This matters because of how Perplexity works. Unlike Claude and ChatGPT, which draw primarily on training data, Perplexity retrieves and cites live web results. It is closer to a search engine than a language model. When a brand drops sharply on Perplexity but holds up on Claude and ChatGPT, it typically signals one thing: strong historical brand authority, but weak current web presence.
In practical terms, this means that what On the Beach built its reputation on. the coverage, content, and citations that trained the earlier models, is not being replenished at the rate needed to maintain visibility on a platform that looks at what the web is saying right now.
The Perplexity test is one of the most useful diagnostics available to travel brands right now. A large gap between your Claude/ChatGPT mentions and your Perplexity mentions tells you whether your AI visibility is built on historical authority or active, ongoing web presence. Only the latter compounds.
Why loveholidays Converts So Well
loveholidays succeeds on AI for a reason that has nothing to do with brand size or marketing spend. It owns a specific, clearly defined position in AI's understanding of the UK travel market: the go-to brand for budget holidays, popular Mediterranean destinations, and value-for-money package deals.
When someone asks which UK travel company is cheapest, loveholidays is the answer. When they ask about booking a holiday to Tenerife, Lanzarote, or the Algarve on a budget, loveholidays is the answer. When they ask which sites offer monthly payment plans, loveholidays is in the answer. AI has learned exactly when loveholidays is the right recommendation, and it makes that recommendation with confidence.
That specificity is an asset. AI systems are not trying to be comprehensive, they are trying to be helpful. They reward brands that they can place confidently in a clear context over brands they can only place vaguely across many contexts.
The Visibility Trap, Defined
The visibility trap is what happens when a brand achieves widespread AI awareness without achieving AI authority. The brand shows up in enough responses to feel well-represented, but it never owns a moment, a specific query type, a specific use case, a specific kind of traveller, where AI says, without hesitation, this one is for you.
On the Beach is the clearest example of this in our data. It is mentioned broadly, described generically, and converted rarely. The brand is in the room, but it is not leading the meeting.
This is commercially significant in a way that simple mention counts obscure. A customer reading an AI response that lists five holiday booking options does not treat them equally. They pay attention to the brand named first, described most specifically, and recommended most confidently. Being third on a list, described as "worth checking," is not the same as being chosen.
What This Means in Practice
The distinction between mention rate and recommendation rate should change how travel brands think about AI visibility. The goal is not to appear in as many responses as possible. The goal is to be the answer to a specific, high-intent question and to be that answer consistently, across all three major platforms.
For brands in On the Beach's position, three things matter most:
- Define a ownable niche. "Beach holidays" is too broad, every major OTA touches beach holidays. The brands that convert well on AI have a sharper identity: budget value, luxury long-haul, late deals, specific destinations. The more precisely AI can place your brand, the more confidently it will recommend you.
- Build current web presence, not just historical authority. The Perplexity drop is a warning. Review volume, recency, and third-party editorial coverage determine whether AI platforms that index the live web continue to surface you. This is not a one-time investment, it requires ongoing output.
- Measure recommendation rate, not just mention rate. If you are only tracking whether AI mentions your brand, you are missing the metric that matters. The question is not whether you appear. It is whether you get chosen.
Frequently Asked Questions
What is AI mention rate vs AI recommendation rate?
AI mention rate is how often a brand appears anywhere in an AI response. AI recommendation rate (or conversion rate) measures how often that mention becomes the platform's top pick. A brand can have a high mention rate and a low recommendation rate if AI knows it exists but doesn't consider it the best answer to any specific question. Our research found that On the Beach converted just 8% of mentions into top picks, while loveholidays converted 70%.
Why do some brands perform better on Perplexity than others?
Perplexity retrieves live web results rather than relying on training data, which means it reflects current web presence rather than historical brand authority. Brands that have maintained high volumes of recent reviews, editorial coverage, and third-party citations tend to perform consistently across all three platforms. Brands that built their reputation earlier but have not kept pace with ongoing content output tend to drop sharply on Perplexity compared to Claude and ChatGPT.
How can travel brands improve their AI recommendation rate?
The most effective approaches are: establishing a clearly defined niche that AI can confidently map to specific query types; building consistent, high-volume review coverage on major platforms; generating regular third-party editorial mentions and citations; and ensuring that brand positioning across owned content is specific rather than generic. AI rewards specificity and confidence. Brands described as "good for comparing deals" will continue to be listed, not chosen.
Is AI visibility the same as SEO?
They overlap but are not the same. Traditional SEO focuses on ranking in search engine results pages. AI visibility is about how language models and AI search platforms represent and recommend your brand in conversational responses. The trust signals that drive AI recommendations, review density, editorial authority, consistent third-party citations, are related to but distinct from the technical and link-based factors that drive traditional search rankings.
What is the Perplexity effect in AI brand research?
The Perplexity effect refers to the pattern (identified in Discovrd's research) where brands that appear frequently on Claude and ChatGPT drop significantly on Perplexity. Because Perplexity draws on live web data, this drop indicates that a brand's AI visibility is based on historical authority rather than an active, current web presence. In our travel study, On the Beach dropped 66% from Claude to Perplexity, the second largest decline of any brand tracked.
