The phone rings in every town you serve

Question

Answer 1

One of six promises

The phone rings in every town you serve

People look in two places now: Google, including the map, and AI assistants they just ask "who can do this near me?" I make sure you're the answer in both, in every town you serve, not only where head office is.

Whatever this is for you, someone is asking for it right now in a town you serve. They type it into Google and look at the map. More and more, they ask ChatGPT or Perplexity and take the first name it gives. If that name isn’t yours, nothing else about your website matters.

Where people look now

Three places, and they overlap less than you’d think. The map, which favours a real location with consistent details and recent reviews. Search results, which favour a clear page for the thing and the place. AI answers, which favour sites a machine can read cleanly and other sites that vouch for you. A business can be strong in one and invisible in the others. Most are.

What being found actually takes

Not tricks. Structure, done early and kept up.

One page per thing you do, and per place you do it. A plumber with eight services in six towns needs the pages to match, not one page that mentions everything.

Your business details the same everywhere. Name, address, phone, hours: identical on the site, the map listing, the directories. Machines treat a mismatch as doubt.

Pages written so an answer can be lifted. The question people ask, answered in the first paragraph, plainly. That’s what gets quoted by an AI and shown by Google.

Fast, clean, readable by crawlers. No heavy builders, no content hidden behind scripts.

Other sites saying your name. Reviews, directories, local mentions. AI answers lean on these heavily.

Built in, then grown

The build gets you found: the structure above is how every site I make is put together, and you launch ranking, not hoping to. The monthly work grows it: new pages for the questions people are asking, listings kept current in every town, AI citations tracked, and a short note each month on what moved.

Proof

For one branch of a portable sanitation group with more than a dozen locations, enquiries from search went from 13 to 28 a month within two months of this work. Visits from AI assistants went from 3 to 17. The rebuilt branch page went from position 14 in Google to the number one AI Overview citation.

More than one location

Head office usually shows up. The branches usually don’t, or share one page and one phone number so nobody local finds them. Each location gets its own page, its own map listing alignment and its own phone, and its calls are counted on their own.

Questions people ask me about this

What is GEO, and what isn’t it? GEO, generative engine optimisation, is making your website easy for AI systems to read, trust and quote, so that when someone asks ChatGPT, Perplexity or Google's AI who to call, you are in the answer. It is not a trick, a plugin or a new kind of keyword. It is clarity, applied to how machines read. Why does GEO matter now? How is search changing? Because the click is disappearing. Google answers more questions on the results page, and a growing number of people ask an assistant instead of searching at all. When the answer is read aloud or shown as a paragraph, there is no list of ten links. There is one answer, and the businesses named in it. What can a website owner control in GEO, and what can’t they? You control what your pages say, how clearly they say it, how consistent your business facts are everywhere, whether AI systems can read you, and what reviews and mentions say about you. You don't control which assistant someone uses, how it ranks sources this month, or whether it cites anyone at all. Work the first list. How do AI systems gather information about a website? Two ways. Some of what a model knows was learned during training, from a snapshot of the web that may be a year or more old. The rest is fetched live: when you ask a question, the assistant searches, reads a handful of pages in seconds, and builds its answer from them. The live route is the one you can influence this month. What’s the difference between ranking and being cited? Ranking is a position in a list; being cited is being used in an answer. A page can rank third and never be cited because nothing in it stands alone as an answer. A page can rank eleventh and be cited every time because its second paragraph answers the question cleanly. The second kind of page is what GEO produces. What makes content quotable to an AI? A passage is quotable when it answers one question completely, makes sense without the rest of the page, states facts rather than implying them, and comes from a page whose purpose and author are obvious. Specific beats general, plain beats clever, and a sentence that could be a caption beats a paragraph that needs context. What is an entity, in plain English? An entity is a thing the machine can point at: your business, a location, a person, a service. Not the words, the thing. "Smith Mechanical" is letters; the entity is the actual company in Barrie with that phone number and those reviews. Much of GEO is making sure the machine connects your words to the right thing. How do we help AI understand who we are? Say it plainly, say it the same way everywhere, and let the machine confirm it. One exact business name. One address and phone per location, identical on your site, Google profile, directories and social accounts. A clear About page. Structured data labelling all of it. Then check how an assistant describes you, and fix what it gets wrong. What is the knowledge-graph mindset? Thinking of your online presence as facts and relationships rather than pages. Business, locations, people, services, and how they connect: this location offers these services, this person leads this team, this business is reviewed here. Machines build exactly that graph about you. The mindset is helping them build the right one. What are the common heading mistakes on a web page? Using headings for size instead of structure. A heading is the outline of the page: one main heading saying what the page is about, then sections beneath it in order. Machines read that outline to understand the page. Big text that isn't a heading, or headings chosen for how they look, breaks it. How do we write self-contained answers? Name the thing, answer the question, give the specifics, stop. A self-contained answer works when pasted into a blank document: no "as mentioned above", no pronouns pointing elsewhere, no dependence on the heading. Forty to eighty words, the subject named in the first sentence, at least one specific fact, and no hedging. Why aren’t we getting more calls? It's nearly always one of two things: people in your area aren't finding you when they look, or they find you and pick someone who looked more reliable. What is the definitional sentence pattern? A sentence shaped like a dictionary entry: "[Thing] is [category] that [does what] for [whom]." It is the single most quoted sentence shape on the web, because it hands an assistant a complete, attributable fact in one line. Every service, location and concept on your site deserves one, and it belongs in the first paragraph. Why don’t our other locations show up? Head office shows up. The branches don't, or they share one page and one phone number, so nobody local finds them. Each place should be found on its own: on the map, in search, in AI answers, with its own phone ringing. How should a page be structured so AI can quote it? Question first, answer immediately, detail after. The heading is the question someone actually asks. The first forty to sixty words answer it completely. The rest of the page earns its place by adding what the first paragraph couldn't. That's how an answer gets lifted whole into an AI reply. How long should a page be? As long as it takes to answer the question, and not a paragraph more. AI systems reward clarity, not word count: a tight 400-word page that answers one thing well gets quoted more than a 4,000-word one that buries it. A page is the right length when every paragraph earns its place. What is structured data, in plain English? Labels machines can read, attached to things on your page: this is our business name, this is the address, this is a service, this is a question and this is its answer. People never see it. Google and AI assistants rely on it to be sure what they're looking at. Which schema types matter most for GEO? Five, for a local service business: LocalBusiness (or a specific subtype) for each location, Service for each thing you do, FAQPage for real question-and-answer content, Person for whoever writes your advice, and Article for anything explanatory with a date. Everything else is nice to have. These five describe the graph from lesson 9. How do we know if our schema is working? Three checks. Validate it with Google's Rich Results Test and the Schema.org validator to catch errors. Confirm Google is reading it in Search Console's enhancements reports. Then the test that matters: ask an assistant to describe your business and see whether the facts it uses are the ones your schema states. What is llms.txt, and do we need one? A plain text file at the root of your site that gives AI systems a short, clean summary of who you are and a list of your most useful pages. It's new and not every assistant reads it yet. It's also cheap to have and does no harm, so sites I build carry one, kept current automatically.