The shortlist is assembled, not retrieved
An assistant answering “who should I call for this” is not looking up a ranked list. It is assembling an answer from sources it has learned to treat as reliable for that category and that place, your site, directories, review platforms, forums, local press, and whatever else it consults.
That is a different job from ranking pages, and it has a different failure mode. A business can rank well and never be named, because the sources feeding the answer never mention it.
Verification comes before quality
The first question is not whether you are good. It is whether the system can establish that you exist, that you are the business the asker means, and that the facts it holds about you agree with each other.
Where those facts conflict, a different suite number here, an old phone number there, a trading name that does not match the licence record, the safest thing for the engine to do is name someone else. Conflicting facts are not a small problem to a system whose whole job is avoiding saying something wrong.
Then extraction
Having decided you are real and relevant, the system needs an answer it can lift. That means a direct statement near a matching heading, not a claim buried in the fourth paragraph of a page about something else.
This is why answer-first formatting is a structural requirement rather than a style preference. Whatever is the clearest extractable statement on your page is what gets quoted, including, if you are careless, a competitor objection you were in the middle of handling.
And it is unstable
These systems are non-deterministic. Around one in ten surfaced domains differs between identical runs minutes apart, so a single check tells you little and a screenshot proves almost nothing.
Any honest approach to measuring this has to run repeatedly and say so. Ours runs each question three times per surface and labels single passes as snapshots.
Read in order, or jump
Twelve chapters, in the order the work has to happen. The first four are the spine; the platform chapters at the end include two surfaces we deliberately do not score, and say why.
01 How AI search picks local businesses 02 Eligibility: can AI even see your site? 03 Robots.txt: block AI crawlers or allow them? 04 Measuring AI visibility honestly 05 Entity foundations for local businesses 06 Reviews as AI input 07 Citation surfaces: where AI engines look 08 Getting ChatGPT to recommend your business 09 Showing up in Perplexity 10 Showing up in Google AI Overviews 11 Showing up in Gemini 12 Copilot, and why we don't sample itSee where you actually stand
A free Visibility Check runs the measurement described in this guide on your own business.
Get my free Visibility Check