What your six-page Discovery report looks like.

AI DISCOVERY AUDIT · MEASURED BASELINE

What is the hotel’s headline discovery condition?

In this audit, your hotel was recommended in 11 of 22 answers to general hotel questions across the two tested engines. FRAGILE

Recognised in 17 of 18 answers when asked about by name.

Two AI engines · 11 general hotel questions · 9 questions naming the hotel · 40 answers tested

Lithari Cove Hotel’s AI discovery baseline is Fragile.

PREPARED FOR

Lithari Cove Hotel
Paphos, Cyprus

DIAGNOSTIC DATE

Illustrative

ENGINES TESTED

Google Gemini · OpenAI

CORE MEASUREMENT

20 prompts · 40 runs / 40 usable responses

How serious is the hotel’s measured position?

Section 01 — Executive Summary

34/100 FRAGILE

Fragile is the lowest of four diagnostic bands: Fragile, Mixed, Advancing and Strong. It is a diagnostic band, not a market percentile or industry benchmark.

Recognised when asked about by name: 17 of 18 answers. Recognition is reported separately and does not enter the Index.

Did your hotel appear among the options?

Section 02 — AI Discovery Position

3 Top Picks · 8 In List · 2 Passing Mentions · 0 Rejected · 9 Absent

Across 22 answers to general hotel questions with a successful response.

ONE TESTED EXAMPLE “Which five hotels would you recommend in Paphos, and why?” · Google Gemini: hotel detected: Yes · Position: In List · OpenAI: hotel detected: No · Position: Absent.

Which websites appear in the answers?

Section 03 — Sources Referenced by AI

Official website 12.5% · Booking platforms 47.9% · Other 39.6%

Across the 96 categorised source references in general hotel answers, 12.5% were the hotel’s official website, 47.9% were booking platforms and 39.6% were other websites. This is not the hotel’s booking-channel mix, and a source reference does not establish a booking, traffic, revenue or why the AI answer was produced.

What should management check next?

Section 04 — Management Priorities

1. 01 — Recommendation coverageWhat we observed: Your hotel was recommended in 11 of 22 answers to general hotel questions. It was absent from 9 of them — 3 from Google Gemini, 6 from OpenAI. When AI was asked about the hotel by name, it recognised it in 17 of 18 answers.What to check next: Identify the traveller questions in which your hotel does not appear, and compare the hotels that appear instead.
2. 02 — Top Pick positionWhat we observed: Your hotel was identified as a Top Pick in 3 of the 11 general answers that recommended it.What to check next: Check which hotels most often take the first recommendation in these answers, and which traveller needs those answers address.
3. 03 — Your website as a sourceWhat we observed: 12 of 96 classified references were your official website.What to check next: Review whether your official website clearly covers the hotel information relevant to these answers, especially where booking platforms and other websites are being referenced.

The audit records what AI answered; it does not establish why.

How was the evidence produced?

Section 05 — Measurement Methodology & Scope

20 core prompts × 2 AI engines · 40 expected · 40 completed · 40 usable

• Positive recommendation presence · 50%
• Official-website reference share · 30%
• Cross-engine positive agreement · 20%

Bands: Fragile 0–49 · Mixed 50–64 · Advancing 65–79 · Strong 80–100. How the Index is built — not a measure of hotel performance.

Point-in-time baseline: AI answers change. Discovery itself does not include implementation, and repeated measurement does not guarantee improvement.