Make your pages easy for AI assistants to read, trust and quote
When someone asks ChatGPT, Perplexity or Google's AI answers what a service costs near them, the assistant quotes a source. Nobody can promise it will be you. What we can do is publish every page in the form assistants read most easily, with the signals they use to judge who to trust, and keep those pages open to the crawlers that fetch them.
What an assistant looks for in a source
An assistant answering a question needs a sentence it can lift, a reason to believe it, and a page it is allowed to fetch. Most small-business websites fail at least one: the price is buried in a PDF, nobody is named as responsible for the content, or the site was never built to be read by a machine.
Every page we publish is built for all three, by default, on every plan.
What we put on every page
| Signal | What it is | Why it matters |
|---|---|---|
| Answer first | One or two sentences that state the fact (the service, the place, the price) before anything else, marked as the page's speakable summary | It is the unit an assistant can quote whole |
| Structured data | Machine-readable JSON-LD naming the author, the reviewer and their credential, your business as publisher, the FAQ, breadcrumbs and the last-modified date | Assistants and search engines read who stands behind the page without guessing |
| Cited sources | Every page lists where its facts come from, with links: your own published information or a public dataset | A page that cites nothing never publishes |
| A Markdown copy | The same page as plain text at the same address with .md on the end, announced in the page's head | Plain text is the format assistants prefer to ingest |
| llms.txt | An index of your pages for AI assistants at /llms.txt, with a fuller version at /llms-full.txt | One file that tells an assistant what your site covers |
Which crawlers get in: your choice, written into robots.txt
There are two kinds of AI crawler. Search crawlers (such as OAI-SearchBot, Claude-SearchBot and PerplexityBot) fetch a page to answer a question and can cite it. Training crawlers (such as GPTBot, ClaudeBot, Google-Extended and CCBot) collect pages to train models, with no link back.
We allow both by default. If you would rather your pages not be used for training, we block the training crawlers in your robots.txt and keep the search crawlers open, because blocking those would defeat the purpose. Googlebot and other search engines are never blocked.
Bing, IndexNow and getting seen sooner
Every page that publishes, changes or retires is announced through IndexNow within seconds, which Bing and several other search engines use. Withheld pages are never announced. Google does not use IndexNow, so for Google we submit your sitemap in Search Console instead.
Pages are rendered on the server, so a crawler reads the full content on its first visit without running any JavaScript.
Limits, and how it is measured
No one controls which source an assistant quotes, and there is no equivalent of Search Console that reports AI citations. We won't send you a citation count we can't verify. What the monthly report does show is visits and customer contacts on your pages, counted by us, and Google's impressions and clicks from Search Console.
AI visibility is not a separate product with a separate fee: the signals above are part of every page we build.
What people ask us
Can you get us quoted by ChatGPT?
No one can promise that. We make it more likely: a quotable answer, named people behind the page, cited sources, a plain-text copy, and the AI search crawlers allowed in.
Should we block AI crawlers?
Block training crawlers if you prefer; it is your call and we set it for you. We advise keeping AI search crawlers allowed, because they are the ones that fetch a page in order to cite it.
What is the Markdown version for?
It is the same page as plain text, which is the format AI assistants read most easily. Open any of our pages and add .md to the address to see what an assistant receives.
Does this replace normal SEO?
No, it is the same work. The answer-first layout, structured data and sources that help assistants are what Google's own results use too.
See what an assistant would read
We build a few free sample pages from your public website, each with its Markdown copy, and send you the links.