Being cited by ChatGPT — appearing as a linked source under an answer — is becoming the new being-ranked-by-Google. Here is what we recommend in 2026, in priority order, with the reasons.

1. Get into the Bing index

ChatGPT search is widely reported to draw partly on Bing’s index. OpenAI’s own crawler docs name OAI-SearchBot as the crawler that surfaces sites in ChatGPT search (step 2), and they do not describe the Bing link. Being indexed by Bing is cheap insurance either way, so make sure Bing has indexed your pages.

Create + verify a free Bing Webmaster Tools account and submit your sitemap.
Use IndexNow (Bing honors it) to push new + changed URLs instantly.
Check coverage in Bing Webmaster Tools.

Many SEO-first guides skip this step because they assume Google indexation is enough.

2. Let AI retrieval crawlers reach your site

AI engines split crawlers by purpose. Retrieval/search bots — OAI-SearchBot, ChatGPT-User, PerplexityBot, Claude-SearchBot — fetch or index pages so they can be surfaced and cited in answers. If your robots.txt or your edge (Cloudflare WAF, a 403) blocks them, you drop out of those answers: OpenAI says sites opted out of OAI-SearchBot will not be shown in ChatGPT search answers, no matter how good the content is.

Explicitly Allow: / the retrieval bots in robots.txt.
Confirm your edge isn’t returning 403 to those user-agents.
Render core content in first-paint HTML — not every AI crawler executes JavaScript.

Our free fix generator writes the exact robots.txt AI-crawler allowlist for you. For the full list of bot tokens, and how to allow search bots while blocking training bots, see the AI bot allowlist guide.

3. Become a recognizable entity

Models attribute facts to entities. If your brand and author aren’t clearly defined, an LLM may credit a competitor for your claim. Make yourself unmistakable:

Organization + Person JSON-LD schema on every page, with sameAs links to ≥2 public profiles.
A real /about page with identity, methodology, and a named author with a byline.
Consistent brand name across <title>, og:site_name, and schema.

The fix generator also outputs this JSON-LD.

4. Write content LLMs can extract

Even with access, you get cited only if your page is easy to quote. Structure for extraction:

Open with a direct, ≤45-word answer to the page’s core question (the part this very page opens with).
Use H2s phrased as the questions people actually ask.
Lead with facts, definitions, and numbers — quote-ready statements, not marketing fluff.
Add a clear published / updated date — freshness can matter for time-sensitive questions.

5. Get corroborated

A fact that appears in several independent places is likely to be picked up more readily than one that only lives on your site. Aim for a claim echoed across your site + a relevant Reddit thread + a LinkedIn post + (ideally) Wikipedia. Build the corroboration graph deliberately: publish the fact, then reference it where your audience already discusses the topic.

6. Measure it — then close the gaps

You can’t improve what you don’t track. Check whether ChatGPT, Claude, Perplexity, and Gemini actually cite you for your target questions, and watch it over time. Start with a free readiness audit (six technical + content dimensions), then generate the fixes.

What about llms.txt?

You’ll see a lot of hype about llms.txt. The honest 2026 answer: none of the crawler docs from OpenAI, Anthropic, Perplexity or Google says it is used to choose sources, so treat it as a harmless nice-to-have, not a primary lever. It costs nothing to add a clean one (the fix generator includes it), but don’t expect it to move citations on its own — the levers above are what matter.

Check where you stand — free.

Audit any URL across six readiness dimensions, then generate your robots allowlist + schema. No signup.

Run the free audit →Generate the fixes →

Or skip pasting a URL — email me a free sample AI-visibility report instead.

How to get cited by ChatGPT.