What answer engines look for before they'll cite you.

• Every page answers one clear question in its first two sentences, not buried in paragraph three.
• Headings are phrased as the actual questions a person would type or ask, not generic section labels.
• Claims are attributed to a specific source or dataset inline, not left as an unsourced assertion.
• The page states a clear, quotable answer in one sentence an AI could lift directly.

• FAQPage or Article schema is present and validates against the required fields, not just decorative JSON-LD.
• The author is a named person with a bio, not "Admin" or the company name alone.
• Publish and last-updated dates are both marked up and visible on the page, not just in the sitemap.
• robots.txt does not block the AI crawlers you actually want reading you, even if you're blocking the ones you don't.

• The page is linked to from at least one other page on the same site that already ranks or gets cited.
• There's a real, checkable fact on the page — a number, a date, a name — not just opinion dressed as insight.
• The content has been live long enough to be re-crawled at least once — new pages rarely get cited in the first weeks.
• The same core claim appears consistently across your site, not contradicted on a different page.

• You've actually typed your own brand or product name into ChatGPT, Perplexity and Google's AI Overview and read what came back.
• You've asked the underlying question a customer would ask — not your brand name — and checked whether you show up at all.
• You've written down what the AI got wrong, since that's the actual list of what to fix next, not a generic checklist.

Prompts worth typing yourself: "What is [your brand]?" · "Best [your category] for [your audience]?" · "[Your brand] vs [a named competitor]" · "How does [the problem you solve] actually work?" Run each in ChatGPT, Perplexity and Google's AI Overview separately — they draw on different signals and rarely agree.