AI Search and GEO · How to
How to get cited by Perplexity: bots, rankings and pages
Part of the guide Generative engine optimization (GEO): a practical guide for 2026
To get cited by Perplexity, allow PerplexityBot in robots.txt and in your firewall (Perplexity publishes its IP ranges as JSON), make sure your pages rank in classic search for the short queries behind your buyers' questions, and write passages that each make one specific, checkable claim. Perplexity's citations overlap with Google's top 10 more than other assistants' do, so classic SEO pays off more here.
How Perplexity chooses what to cite
Perplexity searches the web for nearly every question and shows its sources next to the answer, which makes it the most citation-heavy of the big assistants. Its index is built by PerplexityBot, and pages fetched on a user's request come through a second agent, Perplexity-User. How Perplexity fits among the other engines is covered in the generative engine optimization guide.
The key data point for planning: in Ahrefs' study of 15,000 long-tail prompts (August 2025), 28.6% of the links Perplexity cited also ranked in Google's top 10 for the same prompt, against 12% for ChatGPT, Gemini and Copilot. Perplexity behaves more like a search engine than its peers, so your existing rankings are a stronger predictor of being cited there.
Perplexity's two agents
Perplexity documents two agents, and they behave differently:
| Agent | What it does (per Perplexity) | robots.txt | IP list |
|---|---|---|---|
| PerplexityBot | "designed to surface and link websites in search results on Perplexity. It is not used to crawl content for AI foundation models." | Perplexity recommends allowing it | perplexity.ai/perplexitybot.json |
| Perplexity-User | "supports user actions within Perplexity"; not used for crawling or training | "generally ignores robots.txt rules" because a user requested the fetch | perplexity.ai/perplexity-user.json |
Two consequences. Blocking PerplexityBot removes you from Perplexity's index, and since Perplexity says it is not used for model training, you gain nothing on the training side by blocking it. And a robots.txt rule will not reliably stop Perplexity-User; that agent only arrives because someone asked about your page.
Step 1: Allow PerplexityBot in robots.txt
Add an explicit rule, or make sure no wildcard rule blocks it:
User-agent: PerplexityBot
Allow: /
Check the live file at yourdomain.com/robots.txt, not your repo copy. CDNs and security plugins can add rules for AI agents on top of yours.
Step 2: Allow its IP ranges in your firewall
Bot protection is the most common silent blocker. Perplexity's docs tell site owners to "set up automated processes to periodically fetch and update your WAF rules with the latest IP ranges" from its two JSON endpoints. If your firewall only allows a hardcoded list, the ranges will drift and requests will start failing without anyone noticing.
Then test the user agent:
curl -s -o /dev/null -w "%{http_code}\n" \
-A "Mozilla/5.0 (compatible; PerplexityBot/1.0; +https://perplexity.ai/perplexitybot)" \
https://www.example.com/your-best-page/
A 200 means the user agent is not blocked. Because your test does not come from Perplexity's IP ranges, also look for PerplexityBot in your server or CDN logs and check the response codes it actually received. The broader bot checklist is in the technical SEO for AI search checklist.
Step 3: Rank for the queries behind the question
Perplexity turns a question into searches, so work out what those searches are likely to be and rank for them. For a question such as "which project management tool works best for a five person design studio that bills by the hour", plausible searches include "best project management software for design agencies 2026", "project management with time tracking" and specific "X vs Y" queries. These are illustrative, not captured from Perplexity.
Since Perplexity overlaps with Google more than other assistants, the usual SEO work applies directly: a page per query shape, a clear title that matches it, internal links from related pages, and enough depth to be the best result. How query rewriting works across engines is in ChatGPT query fan-out explained.
Step 4: Write passages Perplexity can quote
Perplexity places its citations next to the statements they support, so each passage on your page should carry one clear claim with its evidence. Practical rules:
- Answer the H2 question in the first two or three sentences below it.
- Put the number, unit and date in the same sentence: "The Pro plan costs $29 a month as of November 2026," not "Pricing is listed below."
- Name things. "Works with WordPress, Webflow and Ghost" can be quoted; "integrates with popular platforms" cannot.
- Link your own sources. Pages that show where their numbers come from are easier to trust and to cite.
The full method is in how to write answer-first content.
Step 5: Be present in the communities Perplexity reads
The GEO guide by Nicholas Dulait (November 2026) reports that Reddit remains a large source for Perplexity even after its share of ChatGPT citations fell in 2026. We could not verify a current Perplexity figure, and published shares vary a lot by method and month. If your buyers discuss your category on a forum, take part as yourself: say who you are in the first line, answer the question, and say when your product is the wrong choice. Fake accounts get removed, and so do the mentions they created.
Treat this as optional and slow. One honest, useful reply every few days is enough to learn whether it shows up in your sources.
Step 6: Test Perplexity the right way
Run each buyer prompt three times, logged out, and write down every source URL, not just the answer. Then ask one natural follow-up question, because follow-ups often trigger new searches and new sources. Log whether you were cited, which of your pages, and which competitors appeared. A free log template and method are in track AI citations for free, and tool options are compared in the AI visibility tracking guide.
Perplexity referrals that pass a referrer usually show "perplexity" in the GA4 source, which you can group into an AI channel.
Common Perplexity blockers
| Symptom | Likely cause | Fix |
|---|---|---|
| Never cited, even for your brand name | PerplexityBot blocked in robots.txt or by a CDN AI toggle | Allow it and check the live file |
| Cited for a while, then not at all | WAF list of IPs went stale | Automate updates from perplexitybot.json |
| Cited for old prices | Page not refreshed | Date every price and review every 90 days |
| Competitor's review page cited about you | No clear first-party page answering the query | Publish an honest comparison or FAQ page |
FAQ
Does Perplexity respect robots.txt?
PerplexityBot does, and Perplexity recommends allowing it. Perplexity-User, which fetches pages for a specific user request, "generally ignores robots.txt rules" according to Perplexity's own docs.
Does PerplexityBot use my content for AI training?
Perplexity says PerplexityBot "is not used to crawl content for AI foundation models". Blocking it therefore removes you from Perplexity's search results without changing anything about training.
Is ranking in Google enough to be cited by Perplexity?
It helps more than for other assistants: Ahrefs found 28.6% of Perplexity's cited links ranked in Google's top 10 for the same prompt in August 2025. Most citations still came from outside the top 10, so passage quality and topical fit matter too.
How quickly does Perplexity pick up a new page?
Perplexity does not publish crawl timelines. Make sure the page is linked from your sitemap and from related pages, then test after a week or two.
Sources
- Perplexity: Perplexity crawlers: PerplexityBot and Perplexity-User roles, UA strings, robots.txt behavior, IP JSON and WAF advice, accessed 16 November 2026
- Ahrefs: AI search overlap: 28.6% versus 12% overlap with Google top 10, accessed 16 November 2026
- Nicholas Dulait: The KING of GEO (X post): Reddit still a large Perplexity source, logged-out testing method, accessed 16 November 2026
Related reading
Last updated: November 16, 2026. Spotted an error? Email support@postbing.com and we will fix it. Editorial policy.
Try Postbing free
Postbing reads your site, plans a month of topics from what your customers search for, writes in your voice and publishes to WordPress, Shopify, Webflow or Ghost. Your first article is free.
Try Postbing freeNo card needed. No link exchanges, ever.