What Ahrefs' AI Search Benchmark Means for Destination Marketers

Ahrefs' AI Search Benchmark gives destination marketers a useful map of how AI search works underneath the answer: which sources get cited, how often citations change, how much rankings still matter, and what kinds of pages show up in recommendation queries.

Most destination and hotel marketers are still using the old SEO frame: rankings, traffic, backlinks, content length, and keyword volume. Those still matter. But AI answers are assembling a view of the world from a moving set of sources.

For places, that distinction matters. A traveler does not ask ChatGPT for "page two of Google." They ask where to go, where to stay, what is worth the trip, what is family friendly, what is underrated, what is open in February, and what is better than the obvious place.

That is not just a keyword problem. It is a place visibility problem.

The traffic story is real, but incomplete

Ahrefs found that AI Overviews reduce clicks to top-ranking content by 58% in position one. Even the tenth organic result lost 19.4% of clicks in their analysis.

Ahrefs also found that ChatGPT has meaningful search-like usage, roughly 12% of Google's search volume by one classification, but Google still sends 190 times more traffic to websites across 76,000 sites using Ahrefs Web Analytics. ChatGPT accounted for 0.21% of total traffic in that dataset.

The wrong takeaway is: AI search is already replacing Google traffic.

The better takeaway is: AI search is becoming a decision layer before the click.

That matters in travel because travel decisions have always been shaped before the last click. A traveler may read a list, ask a friend, watch a video, compare neighborhoods, scan hotel reviews, check an events calendar, and then book somewhere days later. The final click gets the attribution. The earlier influence usually does not.

If your destination is missing from "best wine weekend near Seattle," you may never see the lost demand in analytics. If your hotel is not mentioned when someone asks for "quiet boutique hotels near the beach in Santa Barbara," there may be no abandoned session to measure.

Traffic is the visible part. Recommendation is the earlier layer.

AI answers change sources faster than they change their mind

Ahrefs analyzed 43,000 keywords over a month and found that AI Overviews had a 70% chance of changing from one observation to another. Nearly half of the cited sources, 45.5%, were entirely new.

At the same time, the core message usually stayed consistent.

For place marketers, this changes the work. You are not trying to win one static citation forever. You are trying to make the source environment around your place strong enough that AI systems keep reaching the right conclusion, even when the exact citations rotate.

The model needs enough good material around the entity to understand the place from different angles:

• Official pages that clearly describe the place, neighborhoods, seasons, audiences, access, events, and reasons to visit.

• Partner and listing pages that are accurate, current, and specific.

• Third-party coverage that reinforces the right category associations.

• Reviews, social video, and local media that give the place texture.

• Structured data that helps machines parse the entity, but does not try to do the whole job alone.

At Drifter, this is why we separate Answer Share from Authority Share. Answer Share asks where a place shows up across the traveler questions that matter. Authority Share asks which sources shaped those answers.

You need both. A mention without source understanding is hard to fix. A source audit without answer visibility is hard to prioritize.

Ranking still matters, but it is no longer the full map

Ahrefs found that 37.9% of pages cited in Google AI Overviews also rank in the top 10 organic results. Another 31.2% rank between positions 11 and 100. The remaining 31% do not rank in the top 100.

Traditional SEO is still part of AI visibility. Strong pages, clean structure, authority, and indexable content still help. But AI systems are also pulling from sources that are not simply the top organic result for the original query.

Ahrefs suggests fan-out queries may be part of the reason. Instead of answering only the original query, AI systems may break it into related sub-questions and source from that wider field.

Travel is built for fan-out. Ask "best places to visit in Oregon with kids" and the hidden sub-questions may include beaches, rainy-day activities, family hotels, drive times, school breaks, hikes, food, safety, and whether Portland should be included.

This is where many destination and hotel SEO programs are too thin. They optimize for obvious terms, then leave the supporting questions to travel blogs, OTAs, Reddit, YouTube, Tripadvisor, local media, and stale listicles.

More content is not the lesson

Ahrefs analyzed 174,048 pages cited in AI Overviews and found a near-zero correlation between word count and being cited. The average cited page was 1,282 words. More than half of cited pages, 53.4%, were under 1,000 words. Only 16% were over 2,000 words.

The answer is not to publish longer pages. It is to publish pages that resolve the question clearly.

For destinations, a 700-word page that answers "best time to visit," "who this is for," "what neighborhoods matter," "how to get around," and "what to pair it with" may be more useful than a 2,800-word article that reads like a brochure.

For hotels and attractions, the same rule applies. A short, specific page about family suites near an event, pet-friendly stays close to a district, accessibility, hours, nearby itinerary context, or seasonal notes may do more than another generic guide.

AI search does not appear to reward length for its own sake. It rewards extractable usefulness.

The best X finding matters more for travel than it first appears

Ahrefs found that for top-of-funnel ChatGPT queries, "best X" blog lists made up 43.83% of all source URLs across 26,283 source URLs and 750 queries in software, agencies, and products.

That dataset was not about travel, but the behavior maps directly onto travel discovery.

Travel is full of "best X" intent:

• Best beach towns in Southern California for families.

• Best hotels in Lisbon for a romantic trip with couples.

• Best ski resort towns in the Pacific Northwest for beginners.

• Best weekend trips from New York with a car.

• Best neighborhoods to stay in Tokyo for nightlife.

This does not mean DMOs should publish spammy "10 best" content and rank themselves first. It means teams need to take list intent seriously as infrastructure.

For a destination, that might mean official, useful, query-shaped pages around seasons, traveler types, neighborhoods, events, and trip types. Not fake neutrality. Actual guidance.

For a hotel, it might mean knowing which third-party "best hotel" pages appear in AI answers, whether your property is included, what those pages say, and whether your own site supports the same positioning.

For an attraction, it might mean appearing in the right itinerary lists, family lists, rainy-day lists, accessibility lists, and local guides.

The list is not just a content format. It is how people ask for judgment when they do not yet know the category.

Google AI Mode and AI Overviews do not cite the same sources

Ahrefs compared 730,000 response pairs and found that Google AI Mode and AI Overviews reached similar conclusions 86% of the time, but cited almost entirely different sources. URL overlap was just 13.7%.

That is a warning against single-surface thinking. AI visibility is not one ranking surface. ChatGPT, Perplexity, Gemini, AI Mode, and AI Overviews can agree on the answer while relying on different evidence.

For a place marketer, one platform can make you feel safe while another quietly misses you. A destination may appear in one assistant for family travel and disappear in another for outdoor travel. A hotel may show up in AI Overviews for brand-adjacent searches but not in ChatGPT recommendation prompts.

The useful view is by intent cluster, platform, competitor set, and source.

What destination and hotel marketers should do next

The Ahrefs report points to a practical operating model:

• Map the traveler questions that matter: audiences, seasons, trip types, neighborhoods, nearby alternatives, events, budgets, accessibility, and reasons to visit.

• Check whether you appear in the answers. This is the Answer Share layer.

• Inspect the sources behind the answer. This is the Authority Share layer.

• Strengthen the source layer: official pages, partner listings, third-party guides, review surfaces, videos, local media, and structured data.

• Build a weekly fix list. AI visibility should become work the team can approve and ship.

This is the part Drifter is built around: monitoring what AI tells travelers, showing who appears instead, identifying where AI gets its information, and turning that into what to fix next.

Because destination and hotel teams do not need another abstract dashboard. They need to know what to do on Monday.

See how AI search recommends your place

Drifter Currents monitors what ChatGPT, Claude, Gemini, and Perplexity tell travelers about you.

It shows which traveler questions you appear for, who appears instead, where AI gets its information, and what to fix next.

If you want to see the gaps for your own destination or hotel, start with Drifter Currents. We offer a free snapshot.

FAQ: AI search visibility for destinations and hotels

What is AI visibility for destinations?

AI visibility for destinations is the ability to see and improve how AI assistants recommend, describe, compare, and cite a destination across traveler questions. It includes whether the destination appears in answers, which competitors appear instead, and which sources shape the response.

Do Google AI Overviews only cite top-ranking pages?

No. Ahrefs found that 37.9% of cited AI Overview URLs ranked in the top 10, 31.2% ranked between positions 11 and 100, and 31% did not rank in the top 100. SEO still matters, but AI citation behavior is broader than traditional ranking.

Does long content perform better in AI Overviews?

Ahrefs found no meaningful preference for long content in AI Overview citations. In its analysis of 174,048 cited pages, 53.4% were under 1,000 words, 30.6% were between 1,000 and 2,000 words, and 16% were over 2,000 words.

Why do best places and best hotels lists matter for AI search?

List-style content maps closely to how travelers ask AI assistants for recommendations. Ahrefs found that "best X" blog lists were the most common source type in top-of-funnel ChatGPT recommendation queries across software, agencies, and products. Travel has similar recommendation patterns, from "best beach towns" to "best hotels for families."

What should a DMO do to improve AI search visibility?

A DMO should audit traveler-intent questions, measure where the destination appears, inspect which sources shape the answers, strengthen official pages, improve partner and listing data, and build a weekly action plan around the highest-value gaps.

How can hotels improve visibility in ChatGPT and AI search?

Hotels should identify the recommendation prompts that matter, such as neighborhood, audience, occasion, amenity, and event-based queries. Then they should check whether the property appears, which competitors are recommended, which third-party sources are cited, and whether their own site supports the positioning AI should understand.

Is AI search replacing SEO?

No. AI search is expanding the work. Technical SEO, useful pages, authority, links, and structured data still matter. The new layer is source influence: knowing which sources AI uses to form answers and improving the environment around the place.