When someone asks ChatGPT for a firm like yours, does your name come back?
AI visibility is the share of AI-generated answers in your category that mention, cite or link your business. The work of earning it goes by two names, generative engine optimization (GEO) and answer engine optimization (AEO). We measure where you stand across five engines, then change it, and re-measure every month against a dated baseline so you know whether it worked. Monthly plans from $1,500, with a 3-week Fix as the way in. The free check comes first.
There are plenty of AI-visibility trackers now, and they’re cheap next to the cost of the fix. Every one of them measures. Not one of them rewrites a page, fixes your schema, or reconciles what your Google Business Profile says with what your site says.
That gap is the whole service. The fix is content architecture and entity work, which is exactly what we already do. So we sell the fix, and the measurement comes with it.
Search stopped sending people. Assistants started answering.
Six numbers, each from a named source. They are the reason this is a line item and not a trend.
drop in click-through for top-ranking Google content where an AI Overview appears
Ahrefs, February 2026overlap between Google's top-10 results and the sources AI engines actually cite
Ahrefs Brand Radar, August 2025year-over-year growth in traffic from AI platforms, while organic search fell 27%
HubSpot, April 2026higher visibility in generative responses for optimized content. The methods that worked best were citing sources, adding statistics and adding quotations
Princeton, arXiv:2311.09735, November 2023words in an average ChatGPT prompt, against 3.4 words for a typical Google search
Similarweb, 2025 GenAI Landscape reportWe ran it on ourselves first.
Before we changed a word of this site, we ran our own query set across five engines and wrote the results down. The finding that mattered was not the score. It was the classification: the engines had decided what kind of company we were, and they were wrong about it. Nothing on the site said the wrong thing out loud. The machine-readable layer did, and that is the layer they read first.
So we fixed our own entity, restructured every page opening, rebuilt the schema, and dated the before. The 90-day re-run uses the identical question set. Ask on a call and we will put the actual scorecard on screen, ours and yours, side by side.
Ten concrete things, not a strategy deck.
AI crawler audit
Confirm GPTBot, ClaudeBot, PerplexityBot, Google-Extended, CCBot and Bingbot can actually reach your pages. Five minutes, and it can invalidate everything else.
Baseline measurement
A frozen set of 20 buyer questions put to five engines, scored and dated, with the raw answers kept. Where you appear, which local businesses appear instead, and which URLs get cited.
Entity classification check
What the engines currently think your business is. This is the one that surprises people.
Answer-extraction restructuring
Rewriting page openings to answer the query directly and completely, in the first 200 words, because that is what retrieval systems judge relevance on.
Schema implementation
Organization, Person, Service and FAQPage in JSON-LD, on a shared-id graph, so the machine-readable version of your business says what you say.
Statistical enrichment
Putting real, sourced numbers into body copy. Statistics get extracted. Prose gets skipped.
Citation density
Citing outside sources in your own copy. Content that cites gets cited, and almost no services site does it.
Entity consistency pass
Site, Google Business Profile, LinkedIn, directories: identical name, description and category language. Contradictions dilute the entity.
Ecosystem placement
Where your category's answers actually come from. Reddit is one of the most-cited domains in AI answers. A Semrush analysis of more than 150,000 citations from June 2025, charted by Visual Capitalist in August 2025, ranked it first with a citation frequency of 40.1%, and Profound's analysis of 680 million citations (June 2025) ranks it first among Perplexity's and Google AI Overviews' top sources. Most tracking tools don't watch it at all.
Monthly AI Share of Voice reporting
The same question set, re-run monthly, per engine, against the dated baseline. One number that moves for real reasons, and every number in the report is your own measurement with its date.
A monthly plan, with the Fix as the way in.
The free check comes first and is yours to keep. Then the Fix puts the structure in place, and a monthly plan measures it, holds it and keeps moving it. Month to month, and we will tell you which tier fits before you pick one.
Keep it working
Once the fixes are in, somebody has to make sure they hold, because the engines change without telling anyone. Every month we put the same 20 buyer questions to all five engines and send you a short report: where you came up, who came up instead and what changed, with the engines’ own answers attached. Every quarter we re-check the site itself. So, you always know whether the money you spent is still working.
Keep it growing
Everything in Keep it working, and then we fix what the report finds, every month, without you having to ask. We also write and update the parts of your most important pages the engines read first: the opening lines, the FAQ answers and the answers they were missing. For a business in a category where the answers keep moving.
Keep it everywhere
Everything in Keep it growing, across every location you run, or in a category where the same three competitors come back in every answer. We take one client per category in each metro, so we are never working both sides of the same question.
AI Visibility Fix
Scoped before we start, priced before we start, finished on a date. The 90-day re-run on the same question set is included, so you see whether it worked. Most clients start here, and the monthly plan picks up where the Fix leaves off.
- Make sure the AI engines can actually read your site
- Measure where you stand today: 20 buyer questions, five engines, every answer kept
- Find out what the engines think your business is (this is the one that surprises people)
- Rewrite the openings of your most important pages so they answer the question a buyer asked
- Add the behind-the-scenes code that tells the engines who you are and what you do
- Put real, sourced numbers on the page, because numbers get quoted and prose gets skipped
- Cite outside sources in your own copy, so the engines have a reason to cite you
- Make your name and description match everywhere you are listed
What this is not.
Every one of these is a thing somebody else in this category will promise you.
Questions people ask about AI visibility
What is AI visibility?
AI visibility is the share of AI-generated answers in your category that mention, cite or link your business. The work of earning it goes by two names: generative engine optimization (GEO) and answer engine optimization (AEO). In plain English: when someone asks an assistant for a firm like yours, does your name come back?
How is this different from SEO?
SEO is the foundation and it still matters, because Google AI Mode and Copilot draw from their own search indexes. But the overlap between Google's top-10 results and the sources AI engines cite has fallen to about 12%. Ranking page one no longer gets you into the answer, and being in the answer no longer requires page one. They are two different games now. We break SEO, AEO and GEO down layer by layer, with the full measurement method, at /seo-aeo-geo.
How do you measure it?
We use a fixed set of 20 buyer questions, built from your category, your area and the services on your site, put to ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews through each engine's own interface, and scored on whether you appear, whether your site is cited, and which local businesses appear instead. The question set is frozen the day we start and the same set runs every month, so the number moves for real reasons. We call that your AI Share of Voice.
Can you guarantee I will get cited?
No, and anyone who does is selling you something. These systems change their retrieval behaviour without notice. What we can do is fix the things that are within your control and show you the movement month over month against a dated baseline, with the raw answers attached so you can read them yourself.
How long before anything changes?
We look for the first movement from technical and structural fixes at about 30 to 60 days, once the engines have had time to recrawl. Entity and ecosystem work, which is the larger half, runs longer. We re-measure at 90 days against the baseline because that is the first point where the comparison means anything.
Do I need to be a big brand for this to work?
No. Answer engines pull from whoever answered the question best, which is why a well-structured page from a small firm regularly outranks a large competitor's marketing copy. Small operators often have an advantage here that they never had in classic SEO.
Find out what the engines say about you.
The free check asks ChatGPT a few of the questions your buyers type, reads your site, and shows you the answer that matters most, word for word, in a minute or two. No email to see it. If your buyers are not asking assistants about your category, the result will show that too, and we will tell you so.
Would rather talk it through first? Book a call and we will put your result on screen together.