Local and professional service businesses, cited in AI answers.
It is not a ranking position. An answer engine does not return ten blue links — it returns two or three businesses in a sentence or two. Either you are one of them, or you are not in the conversation at all. There is no second page to be on.
Whether you get named comes down to three things: whether your website states facts a machine can extract, whether those facts agree with every directory and profile that lists you, and whether enough independent sources corroborate them.
*Industry estimate for local businesses surfacing in AI-generated recommendations. Your own baseline is measured, not estimated — that is what the audit is for.
Almost every business we audit is doing something right. Being invisible to AI is rarely about quality — it is about four specific, fixable failures.
Traditional SEO put you on page one. Answer engines don't return pages — they return two or three names, and they choose those names on different evidence than a ranking algorithm uses.
Your hours on your website, your hours on Yelp, your hours on Apple Maps. When sources conflict, a model would rather omit you than state something it can't verify.
No structured data, the useful answer buried three paragraphs into marketing copy, and key content assembled by JavaScript after the page loads. There is nothing there to extract.
Answer engines weight independent corroboration heavily. If no directory, review platform, or local publication confirms what you claim, there is no reason to pick you over someone they can verify.
Four tiers, in the order they should be bought. Everyone starts with the audit — it is the only way to know which of the others you actually need.
Find out exactly where you stand today — on every answer engine, against the competitors you actually lose to.
Make your business machine-readable, and make your facts agree everywhere they appear.
Give answer engines independent sources that corroborate what your website claims.
Structure your catalog now, so you're already readable when AI assistants start transacting.
Not sure which tier you need? That is what the audit answers — and we quote the rest only after we've seen the results.
Start with the audit ↗Every stage produces a number you can check. Nothing here depends on you trusting our judgement over your own reporting.
We run a dozen real customer questions against ChatGPT, Gemini, Perplexity, and Google AI Overviews — for your business and for two competitors you name. You get the scorecard either way. You see the gap before you spend anything on fixing it.
Schema markup, name-address-phone consistency across directories, a complete Google Business Profile, and your highest-value pages restructured to answer questions in the first line. This is the work that makes a model able to quote you.
Reviews, directory presence, local press, and content worth citing. Answer engines weight what independent sources say about you far more heavily than what you say about yourself, so this is where durable gains come from.
We re-run the identical prompt set each month and report three numbers in order: citation rate, AI-referral and branded search activity, then customers who told you they found you through AI. Always against the same named competitors.
Impressions, reach, and “AI mentions” are easy to inflate. These three are not, which is why they are the only ones on the report.
How many of your tracked prompts name you, broken down per engine. This is the first thing to move, usually within eight to twelve weeks, and it moves before anything downstream does.
Visits arriving from AI tools, plus people searching your business name directly. This confirms the citations are reaching real customers rather than just existing in a test.
Customers who told you they found you through an AI assistant. It lags the other two by design — and it is the only number that decides whether this was worth paying for.
We spent years building websites that ranked well. Ranking stopped being enough — so we rebuilt the practice around getting named by the assistants your customers now ask first.
Engineers, designers, and operators. The technical work is done by the same people who ran your audit, and every audit is reviewed by a person before it reaches you.
Before selling AI visibility work, we applied the whole checklist to this website.
Structured data on every page. Content written to answer a question in its first sentence. HTML rendered on the server, so a crawler that never runs JavaScript still sees everything. Explicit access granted to GPTBot, PerplexityBot, and ClaudeBot in our robots.txt. A real dateModified on every page, matching the review date printed at the bottom of it.
The page you are reading is the first audit we ran. Every item on the foundation checklist we would hand a client is one we have already done here — you can verify all of it by viewing source.
We are not publishing client citation numbers, because we do not have them yet. Our own before-and-after is being documented now, and pilot case studies follow after that. When there are results worth showing, they will appear here with the prompt set and the dates attached.
Not keywords — full questions, in the words your customers use. Localised to your category and metro before we run them.
Tell us your business and the one or two competitors you lose to most. We'll run the prompt set against all four answer engines and send you the scorecard — including the competitor comparison.
Reviewed and updated
