A fixed research framework.
Traveller profiles, themes, seasons and AI models are part of a structured study. Your destination and its competitors are assessed within the same framework.
AI Visibility Audit · DMOs, cities & national tourism boards
Measure how AI sees your destination.
Our audits are grounded in our academic research into AI visibility in tourism. A fixed methodology, extensive repeated testing and a large dataset show how ChatGPT, Gemini and Claude represent your destination — with competitor benchmarks and clear priorities for your team.
Inside an AI Visibility AuditSee the evidence.
See the evidence.
Understand the decision.
Explore three findings from our anonymised lakeside pilot: how AI describes a destination, how it compares with peers and where the action plan begins.
Visible Tourism / Research notesAnonymised pilot
Language extracted from the pilotA familiar place.
A familiar place.
An indistinct experience.
The lowest-scoring destination received the most named mentions. Yet the language used to describe it was often generic.
“large freshwater”“glass-clear water”“a lake”
Compare the language used for leading destinations
What it says about the leaders
"medieval castle"
"reflective beauty"
"national park"
"historic town"
"few crowds"
The research question
Does AI explain why a traveller should choose this place, or simply describe where it is?
Four comparable lakeside destinationsMost mentioned.
Most mentioned.
Still ranked last.
Destination D scored 37/100 despite having the highest share of official sources in the field: 16.7%. Discovery position and sentiment exposed the gap.
43index points behind the leader
Priorities recommended from the pilot
A starting point for your team.
- 01Make the experience specific
Connect distinctive activities, places and seasons in official content.
- 02Strengthen overlooked themes
Prioritise underrepresented experiences, including nature and eco tourism.
- 03Coordinate and re-measure
Agree responsibilities across the destination, then track the same core questions.
Your full plan names the pages, domains and partners to prioritise, with a sequence your team can work from.
Your audit includes an executive summary, detailed findings and a prioritised strategy.
Explore the methodology →Why now
Know the answer travellers see. Before your next campaign.
Your next campaign starts from a story that Google and AI already tell. Establish a baseline now, so your content, partnerships and budget address the gaps that matter.
Start with a baseline
Understand how you are represented before committing your next campaign budget. A benchmark gives your board evidence for the decisions ahead.
The gap is measurable
Our four-destination benchmark revealed a 43-point gap on a 0–100 scale. The report explains what separates the scores and where to focus your effort.
Give your team a starting point
Your official pages are only part of the picture. Identify the sources, partners and themes that need attention, then act in a clear order.
What sets our audit apartAcademic research.
Academic research.
Applied at scale.
We conduct academic research into how AI represents tourism destinations. Every audit applies our fixed, documented methodology to a large answer dataset, generated through repeated runs of a defined prompt library.
This lets us examine recurring patterns and variation in how AI recommends and describes your destination — and turn those findings into priorities for your team.
Repeated testing at scale.
The same question can produce different answers. Repeated prompt runs let us examine which recommendations recur and where the answers vary.
A large dataset for each audit.
We analyse recommendations, descriptions and recorded sources across the answer set. The findings show patterns across models and travel contexts, with their scope and limitations made clear.
Inside the methodology
Three levels of traveller questions.
Our prompt framework examines general recommendations, specific travel interests and competitive discovery.
LEVEL 1
General queries
Profile-based recommendations — age, group, budget, length of stay — with no thematic filter. What does the AI suggest on its own?
LEVEL 2
Constrained queries
8 travel motivations × 4 seasons — 32 unique prompts — to test how your destination surfaces theme by theme.
LEVEL 3
Comparative & discovery
Competitor recommendations for the same traveller, plus un-named “best in Europe” prompts — does your destination appear on its own?
34,560
answers in the pilot corpusStudy dimensions: 12 traveller profiles, 8 product groups, 4 seasons, 3 AI models and 5 repetitions per prompt. The six-turn conversation structure includes the four seasonal prompts. Each analysis uses the relevant subset of the corpus.
3AI models
8product groups
12traveller profiles
4seasons
5×repetition / prompt
These figures describe the pilot study. Each audit follows the documented methodology; the size of its dataset depends on the agreed study scope.
What you get
A board-ready report you can act on
- Your AI Visibility Score (0–100) — a single, benchmarked score for your destination, set against up to three competitors.
- A full breakdown — exposure, quality of representation, sentiment, source quality and geographic concentration.
- A model-by-model ranking — where you win, where you lose, and how volatile your visibility is.
- The problem map — the key issues holding your score back, evidenced with real examples.
- Prioritised interventions — a concrete action plan: narrative, content & GEO, thematic anchoring, model-specific fixes.
- Executive + full report — a shareable summary for your board and the full findings for your team.
What your report reveals
Not just a score — the reasons behind it
The benchmark figures and language excerpts below come from the anonymised pilot. The interactive map uses a clearly labelled fictional destination to demonstrate spatial analysis. Your own report identifies your destination, findings and sources.
37/ 100
AI Visibility Score
Seen the most — and still chosen last
Last of four peers · 43 points behind the leaderThe audit explains the observed differences behind the score and identifies evidence-based priorities for your team. In this example, framing and discovery were weaker despite high named exposure and the strongest official-source share. Subsequent measurement helps assess how representation changes after the agreed work.
Qualitative indicators; exact values and ranks are labelled.
Named exposure — when you're asked aboutHighest in field
Source control — your own pages16.7% — best in field
Discovery — unprompted recommendations4th of 4
Sentiment — how you're described+0.27 (field 0.41–0.47)
Framing — vivid vs generic wordsGeographic, not experiential
How recommendations differ across AI assistants
The same destination and question can produce different recommendations across three assistants. The report shows where those differences matter.
2Model AStrong, consistent recommendation#2
3Model BNamed, but after two rivals#3
6Model CRarely surfaced for the same prompt#6
This pilot compares answers from ChatGPT, Gemini and Claude. Your audit covers the systems and questions agreed in advance.
Where you show up — by theme and season
A presence map across travel motivations and seasons reveals which themes carry you and which leave you invisible.
SpringSummerAutumnWinter
CultureStrongStrongFairWeak
Food & wineFairStrongStrongWeak
Nature & ecoWeakWeakWeakWeak
FamilyFairFairWeakWeak
RecommendedMentionedAbsent
A real (anonymised) answer, as travellers see it
"For a lakeside break in this region I'd suggest Town X for its historic abbey and Town Y for its thermal spas and vineyards. Your destination is also on the lake and has a few hotels, but it's usually quieter — more of a stopover than a main base."
Your report captures answers like this verbatim, for every question and every model — so you can see exactly how you're being framed, and against whom.
Where the answer comes from
We check which sources AI actually cites about you. In this case the destination had the best source profile in the field — and even so, its own official pages accounted for a minority of recorded citations.
Your own official sources16.7% — best in field
Curated third-party (OTAs, media, wikis)32.7%
Un-curated general web50.6%
Even in the strongest source profile, most recorded citations pointed to third-party pages. The audit identifies priorities for your official content and the wider sources describing your destination.
Inside the audit · Spatial intelligence
Illustrative datasetA destination is more than its headline attraction.
Explore how a visibility audit reveals the places AI names, the clusters it favours and the local assets it overlooks.
Fictional destination
AI mentions per placeFixed scale across all categories
− 0
14080120160
Colour follows the number of mentions.
Select a place to inspect its exact count.
Lake Serenne · Fictional alpine region · illustrative geography
100%
Point colours show exact mention counts; the heat layer smooths their spatial pattern. All places and data are fictional, separate from the real benchmark above.
96places in view
—with no sample mentions
From map to action. Identify overlooked places, then investigate the content, source coverage and partnerships that could improve their visibility.
How an engagement runs
Your completed audit and action plan — within three months
We manage the full research project. Your team does not need to collect data, build prompts or carry out analysis. The completed report and action plan are delivered at project end, within a maximum of three months from the start.
1
Week 1
Scoping call
We agree your objectives, your real competitors, and the questions that matter for your destination — the KPIs your board cares about.
2
Weeks 1–3
We build your prompt library
Our team applies the fixed methodology to your destination and the agreed study scope, covering relevant traveller motivations and seasons. We design and prepare the query set.
3
Weeks 3–7
We build your audit dataset
We run the defined prompt library across the selected AI models with systematic repetition, capturing the answers for analysis. The dataset lets us examine recurring patterns and variation across models and repeated runs.
4
Weeks 7–12
Analysis, report & strategy
We score and benchmark you against competitors, and turn the data into a board-ready report and a named action strategy — the domains to fix and the partners to work with.
Who it's for: DMOs, city tourism departments, regional boards and NTOs — from a single city to a whole country.
FAQ
Common questions
How much does the audit cost?
We quote each audit for the size of your region and the agreed research scope. A city, a tourism region and a national study require different coverage. We confirm the project price before work starts.
How much work does our team need to do?
The audit requires no operational work from your team. Once the engagement is agreed, we handle the prompt library, data collection, analysis, reporting and action plan.
When will we receive the results?
The audit is a project lasting a maximum of three months from the start. You receive the completed findings, executive report and visibility action plan at the end of the project.
Can we buy the audit without ongoing GEO?
Yes. The audit and its action plan are a standalone service. Ongoing GEO is a separate engagement, which you can decide on later. There is no requirement to continue with us to use the findings.
What is an AI visibility audit?
An AI visibility audit measures how a destination appears in a defined set of AI-generated answers, including mentions, recommendations, descriptions and recorded sources. Visible Tourism uses a documented methodology and a 0–100 benchmark index to compare the destination with the agreed competitors.
How is it different from an SEO audit?
An SEO audit examines technical accessibility, content and search performance. An AI visibility audit examines how selected AI systems describe, mention, recommend and cite your destination in their answers. The two provide complementary evidence.
Can the audit be repeated over time?
Yes. Our fixed, documented methodology and defined prompt set allow the study to be repeated. Individual AI answers can change. Repeated testing helps us assess those changes; model updates and measurement conditions need to be considered when comparing results over time.
What is the research status of the methodology?
The methodology is fixed and documented, and our academic research is continuing. A research data article describing the method and a sample dataset is in preparation for Scientific Data.
Can you audit a whole country?
Yes — the same method scales from a city to a region to a national index across regions and themes.