How we measure AI visibility
AI visibility is whether AI answers like ChatGPT name your store when a buyer asks, and MG Lumeo measures it on a fixed method you can inspect.
The method on this page is the one we run on ourselves.
What measuring AI visibility means
Measuring AI visibility counts how often AI answers name your brand, and which sources they pull from, for a fixed set of buyer questions. Run the same questions through ChatGPT, Perplexity AI, and Google AI Overviews, record the mentions and your share of voice against rival names, and you have a baseline.
A mention is one naming of your brand in an AI answer, and share of voice is your share of those mentions, so a store can hold a strong share on one question and none on the next.
See the same measurement run on your store, with your buyer questions and your own baseline.
Request your free AI auditWhat you see in the reporting
-
-
-
-
-
How the measurement works, step by step
- 01Agree the questionsWe agree the buyer questions your customers ask and the AI engines to test them in, so the method is built around your market, not a generic list.
- 02Pull the baselineWe run the prompt set through ChatGPT and Perplexity AI and record your mentions, your share of voice, and the source domains the engines pull from.
- 03Find the earnable sourcesWe identify the sources you can realistically earn and the gaps against rivals, and cross-check the picture against Google AI Overviews for the same questions.
- 04Re-measure and reportWe repeat the run on a set cadence and report the change against your baseline, so every reporting cycle compares like with like.
These four steps are the measurement side of how MG Lumeo delivers every engagement: Discovery, Strategise, Execute, and Optimise.
What you were told about AI search, and what is true
- Every Dubai agency selling AI search can show you how it measures success.Of roughly twenty Dubai agencies claiming these services, effectively none names its measurement tool or publishes a result.
- An agency can promise your brand a place in ChatGPT or Google AI Overviews.No one can. The engine decides what to name, and Google's own guidance says no special schema or AI-only file makes a page appear in its AI features.
- A zero baseline means a weak score.It means an empty mention dataset: the engines returned no mentions at all, which is a starting point, not a verdict.
A 2023 Princeton study of Generative Engine Optimization found that quotations, statistics, and cited sources increased inclusion in AI answers, while keyword stuffing lowered it.
You see the number against your baseline every reporting cycle and decide whether to continue, with no lock-in.
Request your free AI auditWho runs the measurement, and how to check it
- Our own baselineMG Lumeo ran the instrument on itself on 2 August 2026 and was named zero times across twenty ChatGPT test prompts. Across six Dubai buyer questions run the same day, it was named in none.
- A named leadNour A. Alamiri, CEO and Founder, leads a small senior Dubai team, plus partners.
- Real premisesLatifa Tower, Office #201, Sheikh Zayed Road, Dubai.
- Facts onlyNo certifications, no award badges, no borrowed client logos.
The proof we offer is method transparency and a measurement anyone can repeat.
Questions we get asked
Four things: your mentions, your share of voice, the source domains the AI engines pull from, and your presence in Google AI Overviews. Mentions count how many times an AI answer names your brand across a fixed prompt set. Share of voice is your named share against the other brands the engine returns for the same question. Source domains are the sites an engine pulls from when it answers, and Reddit is the single largest input, so the map tells you what to earn rather than guess.
We check ChatGPT and Perplexity AI, cross-checked against Google AI Overviews for the same questions. The set is agreed with you at the start, so the measurement matches where your buyers ask.
We re-measure on a set cadence and compare each run to your baseline. Fixing the cadence is what makes the change trustworthy, because every run tests the same prompt set.
No. AI visibility and Google ranking are different populations: a store can rank well in Google organic results and still be named by no AI answer. That is why we measure your mentions and your share of voice in the AI engines directly, rather than reading them off your search rankings.
A zero baseline means the engines are not naming you yet, which is a starting point, not a verdict: it is an empty mention dataset, the same result we published for ourselves. The twenty-prompt figure is the global ChatGPT picture on a defined prompt set and date, not a UAE-localised share of voice, because the mention endpoint resolves to a United States location. From there, the source domains show where citations are earned, and the next run measures whether the number moved.
No, and no one honestly can, because the engine decides what to name. We commit to the method, the measurement and the reporting, and we show the number either way.
See your own AI-visibility baseline
Your store has an AI-visibility baseline right now, and the AI-shopping-readiness audit shows you what it is, before anyone pitches you. It is the same instrument MG Lumeo runs on itself.
Request your free AI audit