Open ChatGPT and ask it for the best roofer, or plumber, or electrician in your town. It will not hand you ten blue links the way Google used to. It will name a business or two, in a sentence, the way a friend gives a recommendation. So the question every local business should be asking now is a simple one: when the AI answers that question, where does it get the names, and how does a page become one of them?
The short version is that AI engines do not rank pages, they quote them. They read the web, pull out passages they trust, and stitch them into an answer. The goal is no longer to climb a list. It is to write a page that is easy to lift and safe to repeat. Here is how to do that for a service page.
Answer the question in the first two sentences
AI pulls self-contained chunks. If the answer to “do they do emergency repairs in Spokane” is buried in paragraph six behind a wall of throat-clearing, the model has to dig for it, and it usually just moves on to a page that said it plainly up top. Lead with the answer. Say what you do, where you do it, and who it is for in the first breath. Then explain.
Write one idea per paragraph, with real headings
Models read structure. A page with clear headings and short, single-idea paragraphs is easy to parse, and the easy-to-parse page is the one that gets quoted. A wall of text is not. Use headings that sound like the questions people actually ask, then answer each one directly underneath. You are basically handing the AI its quotes, pre-cut.
Anchor every claim to a real place and a real job
This is where most service pages go limp, and where a local business can win. “We provide quality roofing services” could be anyone, anywhere, so the AI has no reason to repeat it. “We replaced 14 hail-damaged roofs in the Spokane Valley last spring” is specific, checkable, and tied to a place. Models prefer concrete, unique detail over generic copy, because specifics are safer to repeat. Name the neighborhoods. Describe actual jobs. Real work is the most quotable content you have, which is the whole idea behind the check-in effect.
Add the schema that matches what is on the page
Schema is a small block of code that tells a machine what your page is: this is a local business, here is the service, here is the area it covers, here are the common questions and their answers. Use JSON-LD, the format the search and AI crawlers read most easily. One rule matters above the rest: only mark up what is actually visible on the page. Models can tell when the code and the content disagree, and a mismatch quietly costs you trust instead of earning it. Clean FAQ blocks and service details are among the formats AI surfaces most.
Make it obviously, consistently yours
AI leans toward citing a page it can pin to a real, consistent business. Same name, same address, same phone, everywhere it shows up. A clear business identity, tied to your real profiles, tells the model you are a known entity and not a page that appeared yesterday. Consistency is boring, and it is exactly what earns the citation.
Do it yourself, or let GeoUP do it
Every move above is doable by hand. You can write each service page answer-first, describe real jobs by neighborhood, keep your name, address, and phone consistent everywhere, and hand-write clean JSON-LD schema for every page. It works. It is also steady, ongoing effort, and the schema is where most people quietly give up.
That is exactly what GeoUP was built to do. It takes the real jobs a business already completes and turns them into geo-anchored, schema-structured content across the website, the Google Business Profile, and social, on its own. It produces the specific, location-tied, machine-readable pages this whole post is about, straight from real work, without anyone hand-coding a line. So do it yourself, or use the GeoUP app.
The takeaway
Writing for AI is not some new dark art. It is the old advice taken seriously: be clear, be specific, be honest about what you do and where you do it, and back it with real work. Pages built that way were always good for humans. Now they are also the ones the machine repeats out loud. For the deeper background on why AI singles out a few local businesses and skips the rest, we got into it in why ChatGPT cites some local businesses far more than Google does.