March 23, 202614 min read

Travel Brand Visibility Rating in AI — March 2026

Monitoring data from 235 AI responses across 8 providers: Yandex Travel leads in mentions, Aviasales sets records for recommendations (36.6%), and Booking loses due to blocking. Full metrics breakdown.

Владислав Пучков

Основатель GEO Scout, эксперт по GEO-оптимизации

Travel is one of the most competitive niches in AI search. When someone asks a neural network "where to buy cheap flights" or "which hotel booking service is best," the AI returns a list of 3-7 brands. If your brand is not on that list — the customer goes to a competitor without even learning about you.

At GEO Scout, we daily monitor brand visibility across 9 AI providers. On March 23, 2026, we collected a snapshot of the travel niche: 8 AI providers, 235 responses, dozens of brands. Here is what we found.

Methodology: how we calculated

Before diving into the numbers, it is important to understand what exactly we measure.

AI providers in the study: ChatGPT, Claude, Gemini, Perplexity, Grok, DeepSeek, Google AI Mode, YandexGPT (Yandex with Alice).

Sample size: 235 responses to a set of travel queries — from commercial ("best service for buying flights") to informational ("how to plan a budget trip to Europe").

Key metrics:

All metrics are calculated automatically by the GEO Scout platform. An interactive dashboard with provider filters is available in the rating.

Top 10 travel brands in AI

1. Yandex Travel — leader in mentions

Yandex Travel appears in nearly 4 out of 5 neural network responses to travel queries. Average position 2.90 — consistently in the top three.

But there are two nuances worth noting.

First, the recommendation rate is only 0.85%. Neural networks mention Yandex Travel as one of the options, but almost never recommend it explicitly. "You can check Yandex Travel" — that is a mention. "I recommend Yandex Travel for finding hotels" — that is a recommendation. The difference is colossal from a conversion perspective.

Second, Claude shows an anomalously high domain citation — 62.1%. That means in more than 6 out of 10 responses, Claude references travel.yandex.ru as a source of information. This is one of the highest domain citation figures we have seen in any niche. Claude clearly considers the Yandex Travel domain to be an authoritative source of pricing and route data.

2. Aviasales — absolute recommendation record

Aviasales is the main phenomenon of this rating. With a mention rate nearly identical to Yandex Travel (77% vs 78%), Aviasales receives a record recommendation rate — 36.60%. This means that more than a third of all neural network responses that mention Aviasales contain a direct recommendation.

For context: in the hosting niche, the best recommendation rate is around 24%. Aviasales surpasses this level by one and a half times. Neural networks in the travel niche recommend significantly more boldly than in technical niches.

The breakdown by providers shows an even more interesting picture:

• Grok recommends Aviasales in 70% of cases — it almost looks like a partnership integration (although it is not)
• Gemini shows positive sentiment of 76.7% — one of the highest figures
• Domain citation 28.94% — AI models regularly reference aviasales.ru

Why is Aviasales recommended so actively? Several hypotheses:

1. Metasearch engine, not a seller. AI perceives Aviasales as a neutral comparison tool, not a party to the deal. Recommending an aggregator is psychologically easier than recommending a specific airline.
2. Content marketing. Aviasales has invested in its blog, guides, and destination reviews for years. This content has made it into AI model training data.
3. International presence. Unlike Yandex Travel, Aviasales is known outside Russia, which expands the base of sources for AI.

3. Skyscanner — solid middle ground

Skyscanner is mentioned in every other response. Positive sentiment of 60.80% with minimal negative (2.40%) — AI perceives the brand neutrally-positively. This is a typical position for an international brand without local problems: stable, but no wow effect.

4. Trip.com — betting on citability

Every other response contains a mention of Trip.com. But the key metric is domain citation 20%, and in Perplexity it reaches 40%. Trip.com actively invests in content that AI models consider worthy of citation. For Perplexity, which works with real-time web sources, the Trip.com website is one of the main sources of pricing and route data.

5. Tutu — Russian mid-tier

Tutu appears in responses about half the time, but the average position of 6.08 is the end of the recommendation list. AI mentions Tutu but does not rank it at the top. Domain citation 9.79% — AI rarely references tutu.ru as a source. For a brand that dominates the railway segment, this is a weak indicator: neural networks do not perceive the domain as an authoritative content resource.

6-10. The second half of the rating

Booking.com: how blocking kills AI visibility

Booking.com is one of the most well-known travel brands in the world. But in Russia, it is blocked, and neural networks know this.

Positive at only 8.86% — this is a catastrophic figure for a globally renowned brand. For comparison: Aviasales has 85.64% positive, Yandex Travel — 84.15%.

Negative 5.06% — AI actively warns about accessibility issues. A typical neural network response: "Booking.com is a well-known service, but it is blocked in Russia. I recommend considering alternatives: Ostrovok, Yandex Travel..."

A similar situation with Airbnb (13th place): mention rate only 13.62%, positive 12.24%, negative 12.24% — meaning the share of negative mentions equals the share of positive ones. AI literally says something negative about Airbnb in every eighth response.

Business takeaway: website blocking is not just loss of direct traffic. It is a systematic reduction in AI visibility. Neural networks are trained on current data and understand the context of service availability for the audience. If your service is unavailable in a country — AI will not recommend it, even if the brand is globally strong.

Yandex Travel vs Aviasales: mentions vs recommendations

The two leaders of the rating demonstrate fundamentally different AI visibility strategies.

Yandex Travel is a mention-brand. AI knows about it and names it in lists, ranking it high (position 2.90). But it almost never recommends it explicitly.

Aviasales is a recommendation-brand. AI does not just name it, but advises it as a tool. Domain citation is 10 percentage points higher — neural networks more often reference aviasales.ru as a source.

These are two different archetypes of AI visibility:

• Awareness-brand: high mention rate, high position, but low conversion of mention into recommendation. AI knows about the brand from the Yandex ecosystem, but does not find enough independent confirmation for a recommendation.
• Authority-brand: slightly lower in position, but significantly higher in recommendations and citations. AI finds many independent sources confirming the value of Aviasales and is ready to recommend it.

For GEO strategy, this is a key insight: mention rate without recommendation rate is awareness without conversion. Being mentioned is good. Being recommended brings money.

HowTrip: a small site that outperforms giants

One of the most unexpected results of the rating is HowTrip in 19th place.

HowTrip's mention rate is 3.7 times lower than Tutu's. But domain citation is 2.1 times higher. Neural networks mention Tutu more often, but reference howtrip.ru as an information source significantly more actively.

This shows that for AI models, brand size and content size are different things. HowTrip creates expert travel content: detailed guides, budget calculations, destination reviews. This is exactly the content that ends up in cited AI sources.

Strategic takeaway: small niche sites with expert content can have disproportionately high domain citation. AI values depth and expertise, not brand media weight. This is great news for companies that cannot compete with giants in advertising budgets but are willing to invest in content.

Characteristics of individual AI providers

Each AI provider evaluates travel brands differently. Here are the key differences:

Perplexity: facts and references

Perplexity places Yandex Travel at position 1.33 — almost always first. Trip.com gets a domain citation of 40% — Perplexity actively cites this domain when answering questions about prices and destinations. Perplexity works with real-time web search, so content freshness on the website is critically important for this provider.

Claude: domain citation

Claude shows anomalous domain citation for Yandex Travel — 62.1%. No other provider demonstrates such a high level of citation for a single domain. Claude apparently considers travel.yandex.ru to be one of the most reliable sources of travel data in the Russian-language segment.

Grok: bold recommendations

Grok recommends Aviasales in 70% of cases and shows domain citation for aviasales.ru at 66.7%. Grok is the most "recommendation-oriented" of all providers: it does not just list options but actively advises specific tools. For brands that want to get into recommendations, optimizing for Grok may yield the fastest results.

Gemini: positive sentiment

Gemini demonstrates 76.7% positive sentiment for Aviasales — one of the highest sentiment metrics in the rating. Gemini tends to describe brands in a positive light, focusing on advantages rather than limitations.

Neural networks recommend more boldly in travel than in other niches

An interesting observation from cross-niche analysis. In the hosting niche, the best recommendation rate is around 24%. In travel, Aviasales shows 36.6%.

Why does AI recommend more actively in travel?

1. Metasearch engines are neutral ground. Recommending Aviasales or Skyscanner is like recommending a calculator. AI takes no risk because a metasearch engine is not a party to the deal.
2. Objective criteria. Flight price is a measurable parameter. AI feels more confident recommending tools with objective data than services where quality is subjective.
3. Rich training data. Travel is one of the most discussed topics on the internet. Reviews, ratings, comparisons — all of this builds AI confidence in recommendations.

For travel marketers: your niche already gets a high recommendation rate from AI. It remains to make sure AI recommends you, not your competitor.

What to do: practical recommendations

For leaders (Yandex Travel, Aviasales)

• Yandex Travel: increase recommendation rate. The current 0.85% is a strategic vulnerability. More expert content is needed that AI can cite as an independent source, not part of the Yandex ecosystem.
• Aviasales: maintain positions. 36.6% recommendation rate is a record, but it is easy to lose. It is important to continue investing in content and monitoring that competitors (especially Skyscanner) are not growing their metrics.

For mid-tier (Skyscanner, Trip.com, Tutu)

• Strengthen domain citation. Create content that AI will want to cite: detailed comparisons, unique data, expert reviews.
• Work with specific providers. Each AI provider has its preferences. Monitoring by individual providers (available in GEO Scout) will help find growth points.

For outsiders and niche players

• Learn from HowTrip. A small site with 20.85% domain citation is proof that expert content works. Create guides, calculations, unique data.
• Do not try to compete on mention rate. Instead — maximize domain citation and recommendation rate. It is better to be recommended in 15% of responses than mentioned in 40%.

For blocked services (Booking, Airbnb)

• The situation is difficult but not hopeless. AI knows about the blocks. The only path is to create content that emphasizes alternative access methods or mirrors. But realistically: while the block remains in effect, AI visibility in Russia will stay low.

How to monitor your visibility

All data in this rating was collected by the GEO Scout platform. We daily monitor brand visibility across 9 AI providers and calculate metrics automatically.

What you can do right now:

1. Check your brand — go to geoscout.pro and see how AI providers talk about you
2. View the full rating — an interactive dashboard with provider and metric filters is available in the rating
3. Set up monitoring — the free plan includes 3 prompts across 3 AI providers, which is enough to get started

Rating data is updated monthly. Next issue — April 2026.

Key takeaways

Mention rate and recommendation rate are different universes. Yandex Travel dominates mentions (77.87%), but Aviasales captures recommendations (36.60%). For business, a recommendation is more valuable than a mention.

Blocking = death of AI visibility. Booking (positive 8.86%) and Airbnb (negative = positive) — clear proof. Neural networks factor in service availability.

Small brands can win on citability. HowTrip with 20.85% domain citation outperforms Tutu (9.79%) and many other major players. Expert content compensates for media weight.

Travel is the most "recommendation-friendly" niche. 36.6% recommendation rate for Aviasales — higher than the best figures in hosting (24%). AI actively recommends in travel.

Each provider is a separate ecosystem. Claude cites Yandex Travel in 62.1% of responses. Grok recommends Aviasales in 70%. Perplexity puts Yandex Travel first. There is no universal strategy — monitoring each provider separately is necessary. The Command Center of GEO Scout automatically turns this data into a prioritized plan: what content to create, which platforms to strengthen, and what to fix to grow visibility in each AI provider.