grandbazar.ai AI VISIBILITY INDEX
All indexes Current index
METHODOLOGY · PUBLIC & VERSIONED

How the AI Visibility Index works

VERSION 1.0 · MEASUREMENT PERIOD 2026-07-14 · RELEASE 2026-W29 · PUBLISHER: GRANDBAZAR.AI

A measurement protocol, not an opinion. Live logged runs, frozen raw answers, published scoring. Every number in the Index traces back to a stored AI answer.

01 · WHAT WE MEASURE

When a client types a real buying-intent question into an AI assistant — “best hair transplant clinic in Dubai” — which businesses does the assistant name, how often, and in which position? The Index measures that, and only that.

It is not a survey of client satisfaction, not an audit of service or clinical quality, not a review platform, and not a ranking of outcomes. It measures presence across the AI answer surfaces that now sit ahead of how clients discover a business.

02 · MEASURED, NOT MODELED

Some visibility benchmarks publish modeled estimates from an AI model's own general knowledge, without ever running live queries against the systems they score. This Index does not model.

Every issue is computed from live, logged, timestamped responses collected from the named AI systems during the stated measurement window. Raw responses are frozen in our database at collection time and kept as evidence. Any published figure — any score, any rank, any percentage — traces back to the stored answers that produced it.

03 · THE ENGINES

The composite score is computed from three assistants, from a UAE environment. Each is captured the way a client actually experiences it — which, deliberately, is not the same method for all three:

CHATGPT 0.50

Captured from the real consumer app through a live browser session. We do not use the ChatGPT API here: its output differs structurally from what users see in the app, so the API would misrepresent what a client is actually shown.

PERPLEXITY 0.30

Via the official API, whose output matches what users see in the app.

GEMINI 0.20

Via the official API, whose output matches what users see in the app.

AVI (0–100) = 0.50·ChatGPT + 0.30·Perplexity + 0.20·Gemini

How the composite is weighted

Grand Bazar is an index of commercial AI discovery — how often a business is named when people ask AI assistants which provider to choose. Weights reflect each assistant's role in that specific moment, not its overall market size:
  • ChatGPT — 0.50. The dominant entry point for conversational recommendations and the most-used consumer assistant.
  • Perplexity — 0.30. A purpose-built answer engine that over-indexes on comparison and purchase-intent queries — the exact moment this index measures.
  • Gemini — 0.20. Broad consumer reach through Google's ecosystem, weighted for its growing but largely passive, search-integrated surface.
These are the Grand Bazar methodology weights. They are published, identical for every business measured, and reviewed quarterly as commercial AI usage matures.

Engines not named here are not part of the current methodology. When a new engine is added, this page is updated and the version incremented.

04 · THE QUERIES

Each issue runs a fixed battery of neutral, buyer-intent questions for one procedure in one city — the questions clients actually ask, in English.

20 neutral client queries, each asked multiple times across the three engines, producing 180 sampled AI responses this issue (≈ 9 repeats per query, ≈ 60 responses per engine).
Clean, non-personalized environment with UAE geography: no account history, no prior conversation, no brand named in the query.
Queries never name a business. We measure open recommendation behavior, not brand-validation behavior.

Terminology is strict: a “query” is a question we ask — always 20. A “sampled response” is one AI answer to a query — 180 in total. Both values are pulled live from the issue manifest — prompt_count and valid_observations — never hardcoded.

05 · SCORING

Presence rate is the core signal: the share of observed answers in which a business appears at any position. Mentions are position-weighted — a business named first counts for more than one named last — and averaged across all prompts and repeats, then combined across engines using the weights above into a composite AVI from 0 to 100.

For this issue, scoring is computed over 180 sampled AI responses. The composite is a period benchmark tied to the stated measurement window — it answers “how visible was this business to AI assistants during these dates,” nothing more.

06 · WHY WE SAMPLE, NOT SCREENSHOT

AI answers are probabilistic: the same question asked twice can return a different list. That is why the Index never publishes a single screenshot as proof. Every figure aggregates many observed answers, the count is published with every issue (180 this issue), and per-engine differences stay visible in the data instead of being hidden inside one number.

07 · NO ONE CAN PAY TO CHANGE A RANK
No business can pay to change its rank

Scores are calculated automatically under this methodology. Commercial relationships do not influence scores or positions. No business — and not the publisher — can hand-edit a rank. Disputes are re-run publicly and corrected in the next issue.

08 · WHAT THE INDEX DOES NOT CLAIM
It does not measure or imply service quality, safety, clinical outcomes, pricing fairness, or client satisfaction.
It is not an endorsement, accreditation, certification, or recommendation, and not advice of any kind.
It is not deterministic or permanent. AI systems change their answers; a position in one issue does not predict the next.
It does not cover all AI systems, all languages, or all phrasings — only those defined on this page.
Business descriptions generated by AI systems are outputs of those systems, not statements by the publisher.
09 · CORRECTIONS, RE-MEASUREMENT, AND DELISTING

A verified representative of any listed business may request:

01
Correction

If an entry contains an error (name, attribution, matching), write to marsel@trygrandbazar.com with proof of authority to represent the business.

02
Re-measurement

We re-run the standard battery and publish the refreshed figures with a new window date.

03
Delisting

On formal written request, we remove the business from public Index pages. Delisted businesses may still be counted in aggregate, non-identifying statistics.

We acknowledge every request within 5 business days. Where re-measurement confirms an error caused by our data collection or processing, we correct the entry and publish a dated correction note on the affected page.

10 · CADENCE AND VERSIONING

Every issue and every row carries its measurement window. The Index is re-measured from scratch every Thursday. This page is versioned; any change to engines, prompt counts, weights, or scoring is logged below and increments the version.

CHANGELOG
v1.0 — measurement period 2026-07-14 (release 2026-W29) — current.
Feel free to ask
Just reply to this email: marsel@trygrandbazar.com
About this index. The index is not a recommendation to buy any product or service and does not assess the professional quality of any company's services. It measures brand visibility in AI assistants (ChatGPT, Perplexity, Gemini and others): we take the top user queries in a segment, run them through the AI assistants, and score each brand's average mention rate with a single formula. AI answers can be incomplete, outdated or wrong. The index reflects the outputs of third-party AI systems at the moment of measurement — not our opinion. To dispute a result or request proof (the raw AI answer and quote) — marsel@trygrandbazar.com. Clinic names and trademarks belong to their owners.
© 2026 grandbazar.ai · Not a judgment of service quality · Based on public AI outputs
Questions or corrections: marsel@trygrandbazar.com