FOUNDFIRST. BYAI.
More of your customers now ask an assistant instead of scrolling a page of links. “Who can fix a hot water system in Caringbah tonight?” typed into ChatGPT, asked of Copilot, or answered by Google’s AI Overview above the search results. The assistant names one or two businesses and moves on. This page spells out how those answers get chosen, and exactly what I build into every local page so the business it names can be yours.
Is your site open to AI?
Type your website address and I will check, right now, whether the crawlers behind ChatGPT, Claude and Perplexity are allowed to read it. Plenty of good business sites fail this without knowing, usually because of a hosting default nobody chose.
The check reads your robots.txt and fetches your home page as three AI crawlers. It cannot see a Cloudflare or hosting bot setting, so a site that passes here can still block at the edge. Ten checks an hour, nothing is stored.
How an AI picks the business it names.
No assistant publishes its recipe, but the pattern across them is consistent enough to build for. Four things decide whether your business is even in the running.
It can only use pages it is allowed to read.
An assistant looking for a local tradie reads pages its crawler has fetched, or searches an index on the spot. A page that blocks the crawler is invisible to it, however good it is.
It leans on the big search indexes.
Google's AI Overviews and Gemini draw on Google's index. Microsoft Copilot runs on Bing's. A business that only ever thought about Google can be missing from a large share of AI answers without knowing it.
It prefers answers it can lift whole.
A clear, direct answer to the question, with the local specifics, in a passage that makes sense on its own. Vague pages and walls of sales copy rarely get quoted.
It cross-checks the business.
Name, area and services that agree across the page, its structured data and its listings read as trustworthy. Conflicting details read as risk.
What I build into every local page.
Eight things, done on every page, for Google and the AI assistants at the same time. It is the same build I ran for my own account across five states before offering it to anyone, so you can see it working on a real suburb before you sign anything.
Open to every AI crawler
The page and its hosting are checked so the crawlers behind ChatGPT, Claude, Perplexity, Google and Bing can all read it. Plenty of business sites block them without knowing: I found four sites I look after doing exactly that this week, all because of a hosting setting nobody had chosen.
In both search indexes
Google Search Console and Bing Webmaster Tools are set up at launch with the sitemap submitted, and IndexNow tells Bing about every change the moment it is published. Microsoft Copilot runs on Bing's index and Google's AI Overviews on Google's, so a page missing from either is missing from those answers.
See all eight, with what each one does
Structured data a machine can read
Schema markup for your trade, your services, the suburbs you cover and your common questions: the facts stated once, in the format assistants and search engines read directly instead of guessing from the layout.
Answer-first writing
Each page opens with the direct answer to the question a customer actually asks, then the detail. A passage that still makes sense when it is lifted out on its own is the kind an assistant can quote.
Real local facts
Every suburb page is built from the Australian Bureau of Statistics Census data for that suburb and the neighbouring suburbs from the official boundary map. Specific local detail is what separates a page an assistant trusts from a template with the suburb name swapped in.
A plain-text guide for AI
An llms.txt file at the root of the site that sums up who the business is, what it does and where, in plain text. It is an emerging standard for pointing language models at the right facts, and it costs nothing to get right.
Fast, simple pages
Static pages that load quickly and show their content without any script running, because most AI crawlers do not run JavaScript. If the words only appear after a script runs, many assistants never see them.
One set of business details
Your business name, service area and contact details stated the same way on the page, in the schema and against your Google Business Profile. When the details contradict each other, assistants tend to skip the business rather than guess.
The crawler board.
The eight checks every page passes before it goes live, and the crawlers it lets through.
- 1
Open to AI crawlers
GPTBot, ClaudeBot, PerplexityBot and the rest can read it
- 2
Indexed in Google and Bing
Search Console and Webmaster Tools set up, IndexNow on every change
- 3
Structured data
Trade, services, suburbs and questions a machine can read
- 4
Answer-first writing
The direct answer in the first sentence, then the detail
- 5
Real suburb facts
ABS Census data for the suburb and its neighbours
- 6
llms.txt guide
Who the business is, what it does and where, in plain text
- 7
Fast static pages
No script needed to show the words
- 8
One set of business details
Page, schema and listings all agree
- Googlebot
- Bingbot
- GPTBot
- ClaudeBot
- PerplexityBot
- Applebot
Skipped or quoted.
Two pages for the same trade in the same suburb. Hover or tap a difference to see why an assistant passes over one and quotes the other.
- Blocked to crawlers versus Open to every crawlerA hosting default that says no to GPTBot and friends, versus a page that lets them all in.
- The answer is buried versus The answer is the first sentenceThree paragraphs of sales copy before the question gets answered, versus the answer up front.
- A template with the suburb swapped in versus Census facts for the suburbThe same page with the place name changed, versus real detail about that suburb and its neighbours.
- No structured data versus Structured data for trade, services and suburbsNothing a machine can read, versus the facts stated once in the format assistants read.
- Details that contradict each other versus One set of business detailsA different phone number on the page, the listing and the schema, versus the same details everywhere.
Ask an assistant, get one name.
This is the moment the whole thing is built for. Someone asks, the assistant answers with one or two businesses, and the rest never get the call. Every page is built so the business it names can be yours.
Drawn from a local page · Open to AI crawlers · Indexed in Google and Bing · Facts stated plainly
For retaining walls in Logan, your business covers timber, block and concrete walls across Springwood, Beenleigh and Shailer Park, with a quote form that goes to one local builder.
SEO and AI search are the same job, done properly.
There is a lot of noise about AI replacing search. For a local business it works the other way: the assistants read the search indexes, so a page that is clear, local, fast and properly indexed is the page that ranks on Google and the page an assistant quotes. I do not sell AI as a separate extra. It is built into every page, because doing one without the other leaves work on the table.
At the 90-day review, alongside the enquiries and the Google numbers, I ask the main assistants the questions your customers ask and show you what they say, so you see where you stand in both places.
Straight answers.
- Can you promise my business is named by ChatGPT or Google's AI answer?
- No, and nobody honest can. AI answers change with the person asking, the exact words and the day. What I can promise is the groundwork: the page is open to every AI crawler, indexed in Google and Bing, marked up so a machine can read it, and written so a passage can be quoted on its own. That is what gets a page considered at all.
- Do I still need SEO if customers are asking AI?
- Yes. The AI answers draw on the same search indexes, so the page that ranks is usually the page that gets read and quoted. I do both at once on every page. AI search is not a replacement for SEO, it is a second place the same work pays off.
- Which AI assistants does this cover?
- ChatGPT, Google's AI Overviews and Gemini, Microsoft Copilot, Perplexity and Claude. Each reads the open web differently, so the work is aimed at what they share: pages they are allowed to read, that sit in the big search indexes, and that state the facts plainly.
- What about my main website?
- Part of the free area check is looking at whether your main site is open to AI crawlers and in Bing as well as Google. Plenty are not, usually because of a hosting default nobody switched off. If yours is blocked, I will tell you how to fix it, whether or not you hire me.