To get cited in Perplexity, allow PerplexityBot in robots.txt, because Perplexity’s docs say it is the crawler that surfaces and links sites in its search results. Then make pages easy to quote: a direct answer at the top, current facts with sources, and clear headings. Mentions on other sites decide whether Perplexity names your tool when users ask for recommendations.
Key takeaways
- PerplexityBot surfaces and links websites in Perplexity search. Perplexity says it “is not used to crawl content for AI foundation models”.
- Perplexity-User fetches pages when a person asks a question and generally ignores robots.txt, because a user asked for the page.
- Perplexity publishes IP ranges for both bots, so you can verify real visits. Changes can take up to 24 hours to apply.
- In our scan of 541 AI tool sites, only 0.4% block PerplexityBot. Access is rarely the problem. Being worth quoting is.
- Pages with quotes, statistics and cited sources gained the most visibility in AI answers in the GEO study.
How does Perplexity find and cite pages?
Perplexity runs two declared user agents, documented in its crawler docs:
| User agent | What it does | robots.txt | IP list |
|---|---|---|---|
| PerplexityBot | Crawls the web to “surface and link websites in search results on Perplexity”. Not used to crawl content for AI foundation models | Follows it. Perplexity recommends allowing it so your site can appear in results | perplexitybot.json |
| Perplexity-User | Visits a page when a user asks a question, so Perplexity can answer and link to it. Not used for crawling or training | Generally ignores it, since a user asked for the fetch | perplexity-user.json |
When someone asks Perplexity a question, it searches its index, pulls a handful of pages, writes an answer and shows those pages as numbered sources. To be one of them, your page has to be in the index (PerplexityBot) and be the clearest, most useful match for the question.
How do you allow PerplexityBot in robots.txt?
Add a group for it. If you also want to keep training crawlers out, list them separately, because each crawler follows its own group.
User-agent: PerplexityBot
Allow: /
User-agent: OAI-SearchBot
Allow: /
User-agent: GPTBot
Disallow: /That example lets Perplexity and ChatGPT search in and keeps OpenAI’s training crawler out. Our AI crawlers list has the lines for every other AI bot, and you can test your file with the free AI search readiness checker.
How do you verify that a visit is really PerplexityBot?
Anyone can fake a user agent string, so check two things together, as Perplexity itself recommends for firewall rules:
- The user agent contains PerplexityBot (or Perplexity-User).
- The visiting IP address appears in the matching published list: perplexitybot.json or perplexity-user.json. Refresh your copy of the list from time to time, because ranges change.
One caveat is worth knowing. In August 2025, Cloudflare reported that it saw Perplexity fetching test sites through undeclared user agents and IP addresses after its declared bots were blocked. Perplexity disputed the report. If you block Perplexity and still see suspicious traffic, a firewall rule based on behaviour is more reliable than robots.txt alone. If you want to be cited, this does not matter: just allow the declared bot.
What makes Perplexity cite a page?
Perplexity does not publish a ranking formula, so the honest answer comes from research on AI answers in general and from how Perplexity shows its sources.
- Evidence on the page. In the GEO study (KDD 2024), adding quotations from credible sources raised visibility in AI answers by about 41%, statistics by about 31% and cited sources by about 27%. Keyword stuffing made things worse.
- A direct answer near the top. Perplexity writes short answers and quotes the sentence that best fits. A two or three sentence answer under your heading gives it that sentence.
- Specific, buyer-stage pages. In a Ten Speed study reported by Search Engine Journal across ChatGPT, Perplexity, Claude and Gemini (September 2026), product pages drew 24.1% of citations for late-stage buyer prompts, and comparison pages about 13%.
- Mentions elsewhere. When a user asks “best tool for X”, answers lean on roundups, directories and reviews. Ahrefs found branded web mentions had the strongest correlation with AI Overview visibility, and the same logic applies to any engine that summarises the web.
- Fresh facts. Add an “updated” date to pricing and feature pages and keep it honest. Google’s helpful content guidance warns against changing dates without changing content, and that is good practice for every engine.
Worked example: checking our own site for Perplexity
We ran these checks on aiseogazette.com in November 2026:
| Check | Result | What we did |
|---|---|---|
| robots.txt allows PerplexityBot | Pass. Our file has a PerplexityBot group with Allow: / | Nothing to change |
| Key pages load without a login or script-only content | Pass on guides and tool pages | Kept guides as plain HTML text |
| Direct answer at the top of each guide | Pass. Every guide in this hub opens with a bold two to three sentence answer | Kept it under 70 words |
| Sources linked on the page | Pass. Each guide ends with a numbered source list | Linked the original source, not a roundup |
| Perplexity referrals in GA4 (last 30 days) | None recorded, while ChatGPT did send sessions | Track monthly and test prompts by hand |
The last row is the useful lesson. Allowing the bot and formatting the page are needed, but they do not guarantee citations. Perplexity cites pages that already win for the question, so the next step is to run your buyer questions in Perplexity, see which pages it cites today, and make your page clearly better than those.
Perplexity SEO checklist
| Area | Do this |
|---|---|
| Access | Allow PerplexityBot in robots.txt and check your firewall or CDN is not blocking its published IP ranges |
| Answers | Open each key page with a direct answer of two or three sentences |
| Evidence | Add numbers, short quotes and links to original sources |
| Structure | Use question headings, short paragraphs, tables for comparisons |
| Freshness | Show a real “updated” date on pricing and feature pages |
| Mentions | Get listed on relevant directories and named in roundups. See AI brand mentions |
| Tracking | Run 10 buyer questions in Perplexity each month and record which pages it cites |
Where AI gets it wrong
A common claim in AI-written guides is that one robots.txt line controls everything Perplexity does. It does not. PerplexityBot follows robots.txt, while Perplexity-User generally does not, because a person asked for the page. So blocking PerplexityBot keeps you out of Perplexity’s search index, but a user who pastes your URL into Perplexity can still have it fetched. Decide what you want, then set rules for each bot separately.
Frequently asked questions
Does Perplexity use my content to train its models?
Perplexity says PerplexityBot is not used to crawl content for AI foundation models, and Perplexity-User is not used for crawling or training. Both statements are in its crawler docs.
How long does a robots.txt change take in Perplexity?
Perplexity says changes may take up to 24 hours to take effect.
How do I see Perplexity visits in my server logs?
Search your access logs for the strings PerplexityBot and Perplexity-User, then check each IP against the published lists. In GA4, Perplexity referrals appear with the source perplexity.ai. See our GA4 AI traffic guide.
Is Perplexity SEO different from Google SEO?
The basics are the same: crawlable, useful pages. The difference is that Perplexity shows several sources inline for almost every answer, so a clear quotable sentence and good evidence matter more than a top ranking.
Sources
- Perplexity, Perplexity crawlers, checked 21 November 2026
- Perplexity, PerplexityBot IP ranges and Perplexity-User IP ranges
- Cloudflare (August 2025), Perplexity is using stealth, undeclared crawlers to evade website no-crawl directives
- Aggarwal et al. (KDD 2024), GEO: Generative Engine Optimization
- Search Engine Journal (September 2026), Product pages get 24% of AI citations, Reddit and YouTube 4%
- Ahrefs (May 2025), An analysis of AI Overview brand visibility factors
- AI SEO Gazette (November 2026), AI tool visibility study
Part of our SaaS SEO guide for AI tools. More guides in Tool Owner Resources.
