The short version

From a guest's question to a 90-day plan, in five steps.

How we measure whether AI names your property, what every number means, and what we will never claim.

1

Questions

The questions your guests ask AI, by location, occasion and venue type.

2

Runs

Each question asked on ChatGPT, Gemini, Google AI, Perplexity and Claude, in your market.

3

Reading

Every answer checked for who is named, in what position, how you're described and which sources are cited.

4

Scores

Visibility, position, citation share, sentiment and consistency, combined in the HI Visibility Index.

5

Diagnosis – Plan

Five drivers explain the gap, and a ranked 90-day plan targets it.

Steps 1 to 4 are the audit. Step 5 turns it into the plan we carry out with you, and steps 2 to 4 repeat every month.

Questions

Questions: what your guests actually ask.

We suggest questions written the way real guests ask AI assistants, by city, occasion and venue type. You add, edit or remove them before anything is measured. Each question gets a tier.

TierWhat it coversExample
Must-winYour category in your location: the questions that bring direct bookings.“Best boutique hotel near Chiang Mai's Old City?”
CoreThe wider category and the occasions guests browse in.“Romantic villa in Koh Samui with a private pool?”
FoundationalFacts about you that AI must get right.“What time does the rooftop bar open?”

Foundational questions name your property, so we score them separately. They check accuracy, not visibility.

Runs

Same questions, same conditions, every month.

Platforms

Five AI assistants

ChatGPT, Gemini, Google AI, Perplexity and Claude. Being visible on one doesn't mean you're visible on the others.

Market

Asked as your guests ask

Set to your destination and the markets you sell to. Other languages, from German to Chinese, are available as an add-on.

Evidence

Every answer recorded

The full answer, the sources it cites, a screenshot and the date. You can check every number against the evidence.

The metrics

What each number means.

MetricDefinition
VisibilityAnswers that name you ÷ answers checked × 100, on must-win and core questions.
PositionWhere you appear in the answer's shortlist. Lower is better.
Citation shareCitations to sources you own or control ÷ all citations in the answers × 100.
SentimentHow the answer describes you: positive, neutral or negative, with the wording quoted.
ConsistencyKey facts that match across your website, Google profile, booking platforms and guides ÷ facts checked × 100.
HI Visibility IndexA 0 to 100 score that combines the five drivers, per platform and overall.

Every score shows its base, for example “named in 87 of 120 answers”.

Limits

What we don't claim.

Answers vary.

AI is probabilistic, and the same question can get a different answer tomorrow. That's why we keep asking the same agreed set and read the pattern, not a single answer.

Change takes time.

AI tools re-read their sources on their own schedule. Corrections can be quick, but most venues see movement within 60 to 90 days.

Rankings aren't for sale.

Nobody can buy a place in an AI's answer. We change what the AI reads about you, and show you whether it worked.

Attribution is partial.

Most AI answers never produce a click. We add AI referral data where it exists and suggest a simple “how did you find us” question for the full picture.

Unmeasured isn't zero.

If a platform, market or language wasn't tested, we show it as not measured, never as a score of zero.

We only check what we track.

Scores cover the agreed questions, platforms and competitors. Anything outside that set is outside the report.

Specialist FAQ

Questions specialists ask.

Is this rank tracking for ChatGPT?

Partly. Being named and your position work like rankings, across five platforms. The difference is that we also explain why, through the five drivers, and then do the work to change it.

Who designed the method?

Hospitality operators with more than a decade in the industry, built around how guests actually choose where to stay and dine. You don't need a specialist to read the results: every report ends with clear next steps.

How do you collect the answers?

We ask each platform the agreed questions as a guest would, set to your market, and record the full answer, the sources it cites and a dated screenshot.

How many answers is a score based on?

Every agreed question, on every platform, every month. Each score shows its base, for example “87 of 120”.

Can we choose the competitors?

Yes. We suggest three based on who AI names instead of you, and you confirm or replace them. More can be added.

How long is history kept?

Every month is compared with your audit baseline, so the full history stays in your reports. The reports and screenshots are yours.

Who picks the questions we track?

You do, with us. We suggest questions based on how guests ask AI in your market, and you accept, edit or drop them. Be sceptical of any test where the vendor picks the questions alone.

Is there search volume for AI questions?

No one has real question volumes, because the AI companies don't share them. We prioritise by tier and by the bookings a question is likely to bring, not by an invented number.

Do questions that name our property inflate the score?

No. Visibility is measured on must-win and core questions, which don't name you. Foundational questions that do name you are scored separately, for accuracy.

Introductory call

Start by looking. Decide after.

Book a short call and we'll show you, live, how AI answers questions about you and your city today.

EUROPE & UK VENUESNL +31 6 4466 0330
ASIA-PACIFIC VENUESTH +66 95 630 5756
MIDDLE EAST & GULF VENUESinfo@hospitalityintelligence.io