Monthly reporting on whether ChatGPT, Perplexity, and Google AI currently mention your business for the queries that map to your services.
AI visibility is the measure of whether AI assistants mention your business when a user asks a question your business should answer. It is the AI-search equivalent of a ranking position: instead of asking 'where does my page rank', you ask 'is my business mentioned in the answer, and if so, where'. The two metrics are correlated but not the same. A page that ranks #1 in Google can be invisible to ChatGPT, and a page that does not rank at all can be cited by Perplexity.
Tracking AI visibility is the only honest way to know whether the AI SEO work is paying off. The metric is per-query, per-assistant, per-month, with the source the assistant used to ground its answer.
Run a 30-query test panel that covers your main services, your main towns, and the long-tail questions your customers actually ask. For each query, record whether ChatGPT mentions your business, the position of the mention in the answer, and the source ChatGPT used. The test panel is run monthly and the results are the AI visibility report.
I run the ChatGPT panel as part of every AI visibility engagement. The panel uses a fixed set of 30-50 queries per client, with the same panel run every month so the trend is comparable.
Perplexity is a different assistant with a different source mix. The same 30-query panel is run against Perplexity, with the position of the mention and the source Perplexity cited. Perplexity leans more on the live web than ChatGPT does, so the source list is usually different, and the work to get cited is different.
I run the Perplexity panel in parallel with the ChatGPT panel, in the same monthly report, with a side-by-side view of which queries are cited in which assistant, and by which source.
Google AI Overviews sit on top of the regular Google index, so the tracking is closer to traditional SEO. For each of the 30 queries, record whether an AI Overview appears, whether your business is cited inside the Overview, the position of the citation, and the source Google used. The Google AI Overview panel is run monthly, with the same panel as the ChatGPT and Perplexity panels for direct comparison.
Google AI Overviews change frequently. The panel is run on the same day each month to keep the trend comparable.
AI share-of-voice is the percentage of relevant queries in which your business is mentioned inside the AI answer. It is the AI-search equivalent of a search visibility score. A 0% share of voice means the assistants do not know you exist for those queries. A 100% share of voice means every relevant query cites you. Most established local businesses sit between 5% and 25% for their core services before the GEO work starts.
I report the share-of-voice metric at the top of every AI visibility report, with the trend over the previous 6 months and the per-assistant breakdown.
Monthly is the working rate. The AI-search area moves monthly; a quarterly check is too slow to catch a competitor move, and a weekly check is too noisy to be actionable. Monthly reporting is the right cadence, with the report delivered by the 5th of the following month.
I run the panel on the first working day of each month and deliver the report by the 5th. If a major move happens in the middle of the month, I flag it in the next monthly report rather than as a one-off.
A prompt coverage report is the list of queries that are most likely to drive a recommendation in an AI assistant, with the current coverage (is your business mentioned) and the gap to full coverage. The report is the input to the GEO work: it tells you which queries to optimise for, in which assistant, with which source priority.
I deliver a prompt coverage report as the first deliverable of every GEO engagement, with the 30-50 queries the client should be cited for, the current coverage, and the 90-day plan to close the gaps.
Build the entity layer first, then the content layer, then the authority layer. Most local businesses are invisible to the assistants because the entity layer is missing or inconsistent, so the content and the authority work do not get cited. The fix is a 1-2 week entity layer project, then a 4-8 week content cadence, then a 30-90 day authority build. The first citations usually appear within 60-90 days.
I run the prompt coverage report on day one of the GEO engagement, then re-run the report every quarter to track the share-of-voice trend.
A handful of specialist tools have launched in the last 18 months (Otterly, Profound, the early versions of Semrush's AI pack). I use a combination of those tools and a manual panel run on the actual assistants (ChatGPT, Perplexity, Google AI Overviews). The tools give the trend; the manual panel gives the ground truth. The two together are the report.
I do not rely on a single tool. The manual panel is the source of truth; the tools are the productivity multiplier.
Three numbers at the top, the detail below. The numbers: total mentions across all assistants for the month, the change vs the previous month, the share-of-voice percentage for the core 30 queries. The detail: per-assistant breakdown, per-query breakdown, the sources the assistants cited, the next 90-day action plan.
I deliver the report as a single PDF plus a shared spreadsheet, so the client can drill into the data without a separate dashboard. The report is built to be read in 5 minutes and explored in 30.
I run AI visibility monitoring for established local businesses across the Costa del Sol (Marbella, Estepona, Puerto Banus, San Pedro de Alcantara, Mijas, Fuengirola, Benalmadena, Torremolinos, Malaga) plus international clients. Reply within 24 hours.
Pick the option that fits. I read every message and reply within 24 hours, every working day.
AI visibility is the measure of whether AI assistants mention your business when a user asks a question your business should answer. A page that ranks #1 in Google can be invisible to ChatGPT, and a page that does not rank at all can be cited by Perplexity. Tracking AI visibility is the only honest way to know whether the AI SEO work is paying off.
Run a 30-query test panel that covers your main services, your main towns, and the long-tail questions your customers actually ask. For each query, record whether ChatGPT mentions your business, the position of the mention, and the source ChatGPT used. I run the ChatGPT panel as part of every AI visibility engagement, with a fixed 30-50 query panel per client.
Run the same 30-query panel against Perplexity, with the position of the mention and the source Perplexity cited. Perplexity leans more on the live web than ChatGPT does, so the source list is usually different. I run the Perplexity panel in parallel with the ChatGPT panel in the same monthly report.
For each of the 30 queries, record whether an AI Overview appears, whether your business is cited inside the Overview, the position of the citation, and the source Google used. Google AI Overviews change frequently; the panel is run on the same day each month to keep the trend comparable.
AI share-of-voice is the percentage of relevant queries in which your business is mentioned inside the AI answer. A 0% share of voice means the assistants do not know you exist for those queries. A 100% share of voice means every relevant query cites you. Most established local businesses sit between 5% and 25% for their core services before the GEO work starts.
Monthly is the working rate. The AI-search area moves monthly; quarterly is too slow, weekly is too noisy. I run the panel on the first working day of each month and deliver the report by the 5th.
A prompt coverage report is the list of queries most likely to drive an AI recommendation, with the current coverage and the gap to full coverage. The report is the input to the GEO work: it tells you which queries to optimise for, in which assistant, with which source priority. I deliver it as the first deliverable of every GEO engagement.
Build the entity layer first, then the content layer, then the authority layer. Most local businesses are invisible to the assistants because the entity layer is missing or inconsistent. The fix is a 1-2 week entity layer project, then a 4-8 week content cadence, then a 30-90 day authority build. The first citations usually appear within 60-90 days.
A handful of specialist tools have launched in the last 18 months (Otterly, Profound, Semrush's AI pack). I use a combination of those tools and a manual panel run on the actual assistants. The tools give the trend; the manual panel gives the ground truth. The two together are the report.
Three numbers at the top, the detail below. Total mentions across all assistants for the month, the change vs the previous month, the share-of-voice percentage for the core 30 queries. The detail: per-assistant breakdown, per-query breakdown, the sources the assistants cited, the next 90-day action plan.