allrecipes.com

Open the robots.txt file at the root of your site — your host or CMS usually has a setting for it. Find the rules that block the readers AI answers are drawn from and add the lines shown below, so each one is allowed. Save and publish; the readers pick up the change on their next visit. A page can also keep itself out of an answer: remove nosnippet or max-snippet:0 from its robots meta tag, and data-nosnippet from around its main content. If your home page carries a noindex rule in its head, remove that too — it tells every reader, Google included, to ignore the page. Nothing else in the file needs to change, and no other reader gains access.

User-agent: Claude-SearchBot
Allow: /

User-agent: PerplexityBot
Allow: /

Your robots.txt turns away ClaudeBot, Google-Extended. That keeps your pages out of AI training and it is your choice — it doesn’t stop an AI answer from reading you, and it isn’t counted here.

Each missing page answers one question your market asks that mentions competitors and not you. Take the question, answer it plainly in the first paragraph, then back the answer with facts from your own site — numbers, dates, names, with a link to where each one lives. Publish it at a plain address on your own domain and link to it from a page that already ranks. One page a day is a good pace; one a week still moves the number. ReachKit writes these for subscribers, but nothing here needs a subscription.

AI answers mention and link pages that answer a question directly. Writing for that is sometimes called GEO, generative engine optimisation. For each page of yours that has search value but reads like a brochure: turn its headings into the questions a buyer would ask, put a direct answer of two or three sentences right under each heading, and add evidence — a figure, a date, a source you link to. Keep the page’s address; rewriting in place keeps whatever it already ranks for.