What is this guide for?
This guide is for the owner of an established local business who wants to understand how AI assistants decide which companies to name, and what can be changed about it. It runs as a set of chapters you can read in order or jump between, and all of it is free to read.
Your customers are increasingly asking an assistant instead of typing a search. The question is no longer only whether you rank. It is whether you get named when someone asks who to call.
This guide explains how that actually works, in the order the work has to happen. It is written from the six-layer audit we run on real businesses, and where something is uncertain or unmeasurable it says so rather than filling the gap with confidence.
What will this guide not do?
This guide will not promise that following it produces AI citations, and it will not pretend the tactics are the difficult part. Both limits are stated here at the top rather than left for you to discover halfway down the page.
It will not promise that any of this guarantees citations. These are third-party systems with ranking logic that changes without notice, and anyone telling you otherwise is describing a product rather than a reality.
It also will not pretend the tactics are the hard part. The hard part is that most businesses fail at layer one, eligibility, and never find out.
Which chapters does this guide contain, and in what order?
Every chapter below is a separate page, and each one is linked here. The order here is the order of the work: twelve chapters, the first four structural, the last of them covering two platforms we decline to score and why we decline.
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 itWhich AI surfaces does this guide count, and which does it leave out?
Mindflow’s published protocol names the counted set in one line: “Google AI Overviews, ChatGPT and Perplexity as the counted set, each scored and reported separately.” (Mindflow Marketing, methodology, X/12 Protocol v2.16, last updated 26 July 2026.) Every status in the table below is quoted the same way.
| Surface | Status, quoted verbatim | Source |
|---|---|---|
| ChatGPT | “One of the three surfaces we count” | Chapter 08 of this guide, Getting ChatGPT to recommend your business |
| Google AI Overviews | “The second counted surface” | Chapter 10 of this guide, Showing up in Google AI Overviews |
| Perplexity | “The third counted surface” | Chapter 09 of this guide, Showing up in Perplexity |
| Google AI Mode | “Google AI Mode is spot coverage only and is not part of the counted set.” | Mindflow Marketing, methodology, X/12 Protocol v2.16, last updated 26 July 2026 |
| Gemini | “A surface we watch and deliberately do not score” | Chapter 11 of this guide, Showing up in Gemini |
| Copilot | “Copilot is not sampled.” | Mindflow Marketing, methodology, X/12 Protocol v2.16, last updated 26 July 2026 |
On why the excluded surfaces are named at all rather than quietly dropped, the protocol states its own reason: “A protocol that never names what it excludes is not a protocol.” (Mindflow Marketing, methodology, X/12 Protocol v2.16, last updated 26 July 2026.)
How is any of this measured, and what is not promised?
The protocol behind every number in this guide is published in full, including the conditions each run is executed under and what it deliberately excludes. You can read it before you spend anything, and hold our reporting to it afterwards.
Read the full methodRankings are never guaranteed. Anything we could not trace to a primary source is absent from this page, not estimated. The audit runs the same six layers described on the pricing page, and Share of Answer is scored quarterly.
Fair questions
Four questions come up before anyone starts this work: whether it is just SEO under a new name, how long it takes, whether an owner can do it without an agency, and why a firm that sells the service publishes the method. The short answers are below.
Is this the same as SEO?
It overlaps substantially and is not identical. Much of the foundation is shared: a crawlable, well-structured, factually consistent site helps in both. What differs is that an assistant assembles an answer from many sources rather than ranking pages, so surfaces you don't own matter more.
How long does this take to work?
Typically two to three quarters before AI numbers move meaningfully, because much of it depends on third-party sources updating on their own schedule.
Can I do this myself?
Most of it, yes, particularly the eligibility and entity work, which is the highest-return part and needs care rather than tooling. That's why the method is published.
Why publish the method if you sell the service?
Because the alternative is asking you to trust a number you cannot check, and in a category this full of unverifiable claims, showing the working is the only durable differentiator.
See where you actually stand
A free Visibility Check runs twelve real buying questions for your category across three surfaces, and shows you the raw responses.
Get my free Visibility Check