How to get cited by ChatGPT, Perplexity, and Google AI Overviews

All three run a retrieval step against a search index, read the fetched pages, and synthesise an answer with attributions. Google AI Overviews draws on Google's index and appears above organic results for a growing share of queries; in SerionFlow's category, ten of ten money terms tested in September 2026 returned an AI Overview. ChatGPT with search and Perplexity issue their own retrieval queries, often several per prompt, and cite the pages whose passages best answer them. In every case the crawler must fetch the page, the passage must be in the HTML, and the passage must answer the question plainly.

Check robots.txt for groups blocking OAI-SearchBot, GPTBot, ClaudeBot, PerplexityBot, and Google-Extended, and check your CDN for AI-bot blocking, which Cloudflare can enable by default. Fetch a priority page with each user agent and confirm a 200 with full HTML. SerionFlow's free AI crawler access checker and robots.txt checker do this in a minute. This step alone explains a large share of zero-citation sites.

AI crawlers do not execute JavaScript. Fetch your page with JavaScript disabled and count the words; if it is a fraction of what the browser shows, the engines are reading the fraction. When SerionFlow audited itself in September 2026, its pages served 111 to 229 words of static HTML and were cited zero times across six buyer prompts, while competitors with full-HTML comparison pages were cited repeatedly. The fix was to render full page content, headings, tables, and FAQs server-side before writing anything new.

Take the buyer questions from your prompt set and from conversational queries in Search Console. For each, publish a page whose H1 is the question, whose first paragraph answers it in forty to sixty words with no preamble, followed by a short TL;DR, sections that each open with a direct statement, a table with real figures, and an FAQ with verbatim questions and short answers. This structure is what all three engines lift passages from. SerionFlow generates drafts in this shape for use-case, how-to, comparison, and listicle intents, but the shape is reproducible by hand.

Engines favour sources that add information and that read as balanced. A fact from your own product data or Search Console, dated and explained, gives the page something no competitor page has. A paragraph stating where your product is the wrong choice makes the page read as a source rather than an advertisement, and engines are measurably more willing to cite balanced sources for comparison prompts.

Run a citation gap analysis to see which third-party pages are cited for your prompts, then make sure your product is accurately listed on them: comparison sites, community threads where relevant, documentation hubs, and reviewers. Supply facts, not pitches. This is AEO link building; the goal is accurate presence on cited sources rather than link volume.

Freeze a prompt set, run it monthly across the three engines, log every cited URL, and track your citation share per engine. Pair it with conversational query impressions in Search Console and AI-related query signals from Bing. Expect a trend over three months. When a prompt stays uncited, compare your page with the cited one on the six steps and fix the gap. Nothing in this method can force a citation; it removes the reasons engines have not to cite you.