What an AI citation actually is
A citation is a URL that an AI engine references in its answer. Perplexity makes citations explicit - every claim is footnoted to a source. ChatGPT (with browsing or search) embeds them inline as markdown links. Either way, the cited URL is the one a curious user can click through to verify the claim.
Citations matter because they are the new backlinks. Being cited by Perplexity or ChatGPT for "the best CRM for startups" sends qualified traffic to your page and reinforces your standing the next time the model answers a similar question.
Why getting cited is the goal
- Direct traffic - users who click footnotes are explicitly verifying. They convert at high rates.
- Compounding signal - the citation itself becomes evidence the next time a related query is asked.
- Brand reinforcement - even unread citations attach your domain to authoritative answers.
- Defensive moat - cited brands are harder to dislodge from a category narrative.
How to read your result
- Citation count - how many distinct URLs the model leaned on. More citations = a richer, more verifiable answer.
- Domain coverage - which sites dominate the response. If review sites and competitor blogs dominate, you have a coverage gap.
- Your domain status - if you supplied a brand, we highlight cited URLs from your domain so you can see at a glance whether you made the cut.
- Answer snippet - the first 500 chars of the response. Useful to verify the model gave a substantive answer (and not a refusal or generic disclaimer).
How to earn more AI citations
- Publish single-question pages that answer one query exhaustively. AI engines prefer pages with one clear thesis.
- Use
FAQPageandHowToschema. Both translate cleanly to citation-friendly snippets. - Cite your own sources. Pages with strong outbound citations are themselves cited more often - it is reflexive.
- Get covered on the third-party sites Perplexity already trusts (G2, niche communities, news outlets, Wikipedia where appropriate).
- Improve your GEO score on each page you want cited.
The fastest way to earn AI citations isn't to publish more on your own blog. It's to get cited by the sites the AI already trusts. Find the top three citation sources in your category and concentrate guest posts, expert quotes and product mentions there. Compounding follows.
Citation source archetypes
AI engines pull from a small number of source types repeatedly. Knowing which type you need to be on is half the battle.
| Archetype | Examples | Why AI cites them |
|---|---|---|
| Review platforms | G2, Capterra, TrustRadius, Trustpilot | Aggregated user voice with structured ratings. |
| Comparison roundups | Top-N blog posts, "best X tools" articles | Pre-organised competitive sets. |
| Authoritative news | TechCrunch, The Verge, Wired, industry trade press | High citation density and freshness. |
| Reference works | Wikipedia, official spec sites, RFCs | Definitional anchor points. |
| Forums | Reddit, Hacker News, Stack Overflow, niche Discords (indexed) | Real-user signal at scale. |
| Vendor docs | Your own /docs, /api, /integrations | Authoritative source on the product itself. |
| Aggregators | Producthunt, Indie Hackers, niche directories | Discovery surface for newer brands. |
Common mistakes
- Reading a single result as the truth. Citations vary across runs. Treat the result as one sample, not the canon.
- Fixating on your own blog citations. Self-citations are weaker signal than third-party citations. Diversify.
- Counting citations from any low-quality source as wins. AI engines de-weight thin sites. Prioritise the eight archetypes above.
- Ignoring engine differences. Perplexity citations follow different patterns than ChatGPT citations. Track both separately.