How to get cited by Perplexity in 2026
If your AEO playbook is built for Google and ChatGPT, Perplexity will quietly leave you out. It retrieves in real time, favours community sources over polished brand pages, and runs its own crawler with its own rules. Getting cited there is a different job, and here is how it works.
Perplexity is small in traffic terms - roughly 1.3% of AI referral traffic as of May 2026, per Similarweb's panel estimates - but it matters out of proportion to its size, because its users are researchers and buyers who trust its inline citations. And it decides who to cite in a way that catches brands off guard.
How Perplexity actually finds sources
Two things make Perplexity distinct. First, it runs its own crawler, PerplexityBot, and its own index, rather than leaning entirely on Google or Bing the way some assistants do. Second, its search layer, Sonar, retrieves from the live web on every query, so answers are informed by current sources rather than a static snapshot. Every answer ships with inline citations by design.
The practical consequence: being in Perplexity's index is a prerequisite, and freshness is a real advantage. A page updated last week can be retrieved and cited this week.
Curious how AI engines describe your brand right now? Get a free visibility audit and see where you stand across ChatGPT, Gemini and Perplexity.
The robots.txt detail that trips people up
Perplexity uses two agents, and they behave differently. Its own documentation is unusually candid about this. PerplexityBot, the indexing crawler, respects robots.txt, and Perplexity explicitly recommends allowing it. Perplexity-User, which fetches a page when a live user query needs it, "generally ignores robots.txt rules" because a user requested the fetch - that is a direct quote from their crawler docs.
On top of that, in August 2025 Cloudflare said it had caught Perplexity using an undeclared stealth crawler that spoofed a normal Chrome browser to get around no-crawl directives, and de-listed it as a verified bot. Perplexity called the report a publicity stunt. You do not need to take a side to draw the lesson: if you want Perplexity citations, make sure PerplexityBot is allowed, because blocking it removes you from the index it cites from.
"Perplexity's own docs admit one of its agents generally ignores robots.txt. The other, the one that indexes you, you should allow."
What Perplexity actually cites
This is where Perplexity diverges most sharply from its rivals, and where the data is clearest. In Profound's analysis of 680 million citations across the major engines (August 2024 to June 2025), Reddit was Perplexity's single most-cited domain - the highest concentration on any one source of any engine studied. ChatGPT's top source, by contrast, was Wikipedia.
A separate 2026 study by Otterly.ai found Perplexity cited official brand-owned domains least among the major engines - well below Google AI Overviews. Read together, the picture is consistent: Perplexity prefers community discussion, forums and third-party validation over a brand talking about itself.
One caution worth stating plainly, because the internet is full of confident numbers here: the exact percentages vary wildly between reports, often because they measure different things (share of all citations versus share of the top ten sources). Treat the direction as reliable and any single hard percentage with suspicion.
What to actually do about it
- Allow PerplexityBot. Check your robots.txt and bot rules. This is the one non-negotiable, because everything else depends on being in the index.
- Earn genuine presence in community sources. Since forum and community content is disproportionately what Perplexity cites, being accurately discussed on Reddit and similar Q&A threads matters. Participate genuinely; do not spam, which backfires.
- Publish original data. Novel statistics and proprietary research are high-leverage across every engine, and Perplexity's real-time retrieval means fresh data can surface quickly.
- Do not rely on your own pages alone. Because Perplexity cites brand domains least, third-party corroboration carries more weight here than on Google. Get others describing you accurately.
- Keep it fresh. Update timestamps and refresh key pages, because a live-retrieval engine rewards recency in a way a static index does not.
The takeaway
Perplexity is not a smaller version of Google. It builds its own index, retrieves live, and cites community and third-party sources far more than brand pages. If you have optimised only for Google and ChatGPT, you are optimised for the wrong sources here. Allow its crawler, earn honest third-party validation, publish fresh original data, and measure whether it actually names you - because on Perplexity, the rules are genuinely different.
See if Perplexity names you
Perplexity cites a different web than Google or ChatGPT, so your visibility there is its own number. Stellarcast tracks whether you are named and cited across all the major engines, Perplexity included. Request a free audit.
Get your free visibility auditFrequently asked questions
How do I get cited by Perplexity?
Allow PerplexityBot in your robots.txt (Perplexity explicitly recommends this in its crawler docs), because it is the crawler that builds the index Perplexity cites from. Then focus on the sources Perplexity actually favours: it leans heavily on community and forum content, cites brand-owned pages less than Google does, and retrieves in real time, so recency and genuine third-party validation matter more here than on Google. Publishing original data and being discussed on sites like Reddit both help.
Does Perplexity respect robots.txt?
It depends on the agent. Perplexity's own documentation says PerplexityBot, its indexing crawler, respects robots.txt and should be allowed. But it also states that Perplexity-User, which fetches a page when a live user query needs it, generally ignores robots.txt because a user requested the fetch. Separately, in August 2025 Cloudflare accused Perplexity of using an undeclared stealth crawler to evade no-crawl directives, which Perplexity disputed.
How is Perplexity different from ChatGPT for citations?
They source from different places. Analysis by Profound of 680 million citations found Reddit is Perplexity's single most-cited domain, while ChatGPT's is Wikipedia. Perplexity also cites official brand pages less often than Google AI Overviews does. In short, ChatGPT skews encyclopedic and Perplexity skews toward community and forum content plus fresh, real-time sources.