When someone asks an AI which business to call, we want it to say yours
More customers now start with an assistant instead of a results page. “Who is a good plumber in Puyallup?” gets an answer with two or three names and no scrolling. Being one of those names is a different job from ranking tenth on Google, and it is work you can actually do.
The answer replaced the list
A search results page gives ten links and lets the customer choose. An AI assistant gives one answer, sometimes with a handful of businesses named and a few sources cited. If your business is not in the answer, the customer never sees you — there is no page two to be on.
Assistants build those answers by retrieving pages, profiles, directories and reviews, then summarising what they find. So the question stops being “where do I rank” and becomes “can a machine tell what this business is, where it works, and why it should be trusted — in a form it can quote?”
| Search results page | AI assistant answer | |
|---|---|---|
| What the customer sees | Ten links, ads above them | One answer naming two or three businesses |
| How you get in | Rank on page one | Be retrieved, understood and quotable |
| What decides it | Relevance, links, on-page SEO | Entity clarity, structured answers, corroboration across sources |
| How it is measured | Position by keyword | Whether you are named, and which page gets cited |
Both matter, and the same underlying work feeds them — the difference is what the customer ends up looking at.
Five things that decide whether you get named
A machine-readable identity
One consistent business entity: the same name, address, phone, hours and services on your site, your Google profile and every directory, wired together with Organization and LocalBusiness schema and sameAs links to your real profiles.
Answers, not brochures
Pages that answer the question in the first two sentences, with the heading phrased the way a person asks it. Assistants quote passages, so the passage has to be quotable on its own.
Evidence a model can find
Reviews, citations, local mentions and named clients. Assistants weigh corroboration across sources, so proof that only exists on your own site counts for little.
Specifics over adjectives
Service areas, prices where you publish them, hours, brands you work on, what you do not do. Concrete facts survive summarisation; “award-winning quality service” does not.
Clean technical access
Server-rendered HTML, fast pages, a correct sitemap, an llms.txt where it helps, and no blocking of the crawlers that feed AI answers.
Measurement you can check
A prompt set we run monthly against the assistants your customers use, plus AI referral traffic split out in your analytics.
Where the effort goes in a first engagement
Illustrative — not client dataA typical split for a first three months; yours depends on what is already in place.
The engagement
- Baseline. We run a set of real customer questions through ChatGPT, Google AI Overviews, Perplexity and Gemini, and record who gets named for each one — you, a competitor, a directory or nobody.
- Entity clean-up. Name, address, phone, hours, categories and services made identical everywhere they appear, with schema that ties your site to your profiles.
- Answer pages. The questions that produced no mention become pages written to answer them directly, with FAQ and service schema attached.
- Corroboration. Directory listings, local citations and a review routine, so the claims on your site are backed somewhere else.
- Re-test and report. The same prompts every month, side by side with the baseline, plus AI referral sessions in analytics.
What nobody can sell you
There is no paid placement inside an AI answer, no submission form, and no agency that can guarantee ChatGPT will recommend you. Anyone promising a ranking in an assistant is selling something they do not control.
What is controllable is everything the assistant reads: the clarity of your identity, the quality and structure of your content, and the amount of independent evidence that you exist and do good work. That is the work we do, and we report it against a baseline so you can see whether it moved.
Much of this also improves ordinary search rankings, because the same signals feed both. If you are already doing SEO with us, a large part of this is an extension rather than a second bill.
What the monthly report shows
AI search questions
Is this just SEO with a new name?
It overlaps, and we will tell you where. The differences are real: entity consistency and schema matter more, answer-shaped passages matter more, and success is measured by whether an assistant names you rather than by a position on a results page.
Can you guarantee ChatGPT will recommend my business?
No, and neither can anyone else. We baseline where you stand, do the work that influences it, and re-test monthly so you can judge the change.
Which assistants do you test?
ChatGPT, Google AI Overviews, Perplexity and Gemini as standard. If your customers use something else, tell us and we will include it.
Does this matter for a local service business?
Increasingly. “Near me” questions are exactly the ones people are asking assistants, and the answer usually names two or three businesses. For a local trade that is a shortlist you either make or miss.
What do you need from us to start?
Access to your website and Google Business Profile, your real service list and area, and thirty minutes to tell us the questions customers actually ask you.
Find out what the assistants say about you
Send us your business name and website. We will run a set of real customer questions through the major assistants and send you the answers — who gets named, who gets cited, and where you sit.