To get cited by Perplexity, first make sure PerplexityBot can reach your pages, then publish pages that answer specific questions directly with current, checkable facts. Perplexity searches the web when it answers and shows its sources as numbered citations, so a page that is crawlable, clearly structured and more useful than the alternatives has a real chance of being cited. Being described accurately on the forums, review sites and publications it already cites for your topic improves your odds further.
This guide covers what Perplexity has said publicly about how it finds sources, where the details are unknown, and a sequence of steps you can work through in a few weeks. If you are new to the field, what is answer engine optimization explains the basics first.
Key takeaways
- Access first: PerplexityBot is the crawler that lets Perplexity find and link your pages. If robots.txt, a CDN or a firewall blocks it, you are unlikely to be cited from your own site.
- Answers are live: Perplexity runs web searches as it answers, so fresh, accurate pages can be picked up without waiting for a new model.
- Citations are visible: every cited source is shown as a numbered link, which makes Perplexity one of the easier engines to audit.
- Specific beats general: pages that answer one question clearly, with numbers, dates and sources, are easier to cite than broad overview pages.
- Third-party sources count: Perplexity often cites forums, reviews and publications, so your presence there matters alongside your own site.
How Perplexity finds and cites sources
Perplexity is an answer engine built around search. When you ask a question, it searches the web, reads a set of pages, writes an answer and attaches numbered citations to the sources it used. A list of sources sits with the answer, so readers can see and open every page it relied on.
Its more thorough research modes run more searches and read more pages before answering, which widens the set of sources considered for a single question. Follow-up questions in the same thread trigger fresh searches, so one conversation can cite many different pages.
Perplexity documents two crawlers that matter to site owners:
- PerplexityBot is the crawler Perplexity uses to surface and link websites in its search results. Perplexity says it is not used to crawl content for training AI foundation models.
- Perplexity-User fetches a page when a user’s question needs it. At the time of writing, Perplexity’s documentation describes these fetches as user-initiated and says they generally do not follow robots.txt for that reason.
Perplexity also publishes the IP ranges its crawlers use, so you can check that traffic claiming to be PerplexityBot is genuine.
What is not known
Perplexity has not published how it ranks candidate pages, how many it reads for a typical question, or how it decides which sources earn a visible citation. It runs its own crawler and index and also retrieves live results, but the exact mix, and how it differs between modes and models, is not public.
What you can observe is the output. For any question you care about, you can see which domains Perplexity cites, how often, and which competitors appear. That makes it a good engine to learn from, as long as you look at patterns across many questions and repeated runs rather than at one answer.
Tip: Paid Perplexity plans let users pick different underlying models and modes, and answers can vary between them. When you track citations, keep your setup consistent so that changes reflect your work rather than a different configuration.
Step 1: Let Perplexity’s crawler in
Open your live robots.txt in a browser and check how it treats PerplexityBot. If no group names it, the rules under User-agent: * apply, and if those disallow your content, Perplexity is disallowed too. To allow it explicitly, add a group that starts with User-agent: PerplexityBot and contains Allow: /, plus any folders you genuinely want to keep private.
Then check the layers robots.txt does not control:
- CDN and firewall rules. Some CDNs and security services offer one-click AI bot blocking, and newer setups may switch it on by default. Make sure PerplexityBot is not caught by it.
- Security plugins and rate limits. Aggressive rate limiting can return errors to crawlers even when robots.txt allows them.
- Caches. Make sure the robots.txt that visitors receive is the current one, not a stale cached copy.
The free AI robots.txt generator builds rules for PerplexityBot and other AI crawlers in one pass, and our guide to which AI crawlers to allow explains how to decide between training, search and user-triggered bots.
Step 2: Find the questions your buyers ask Perplexity
People often use Perplexity for research: comparing options, checking facts, understanding a topic, finding the right tool for a job. Build a list of 25 to 50 questions that match that intent and lead to a decision in your market. The free AI prompt ideas generator gives you a starting list grouped by stage.
Ask each question in Perplexity and record three things: whether you are cited, which domains are cited instead, and which kinds of source dominate. You will usually see a mix of vendor sites, review platforms, forums, video, news and reference sites, and the mix differs by topic. That mix is your map. It tells you whether to win on your own site, on third-party sources, or both.
Perplexity also suggests related follow-up questions under many answers. Note them down. They show where the conversation tends to go next, and each one is a question you can answer on your own site.
Step 3: Publish pages that answer one question well
Perplexity reads pages to find the part that answers a specific question. Broad pages that cover everything lightly are easy to skip. Focused pages, or clearly separated sections, are easy to cite.
- Answer in the first lines. Under the title and under each key heading, give a direct two or three sentence answer before the context.
- Match the question in the heading. If buyers ask “how much does X cost for a team of ten”, have a heading that says so and an answer right below it.
- Make every section self-contained. A passage quoted on its own should still make sense, so name the subject instead of writing “it” or “this tool”.
- Use lists for steps and options. Numbered steps, criteria, pros and cons are easy to extract without losing meaning.
- Cut the preamble. Long introductions push the answer down the page and dilute the passage that matters.
Before publishing, run the draft through the free answer readiness checker. It flags buried answers, vague headings and missing specifics, the problems that most often keep a good page from being quoted.
What a citable answer looks like
It helps to see the difference side by side. Imagine a buyer asks how long it takes to set up your product.
- Hard to cite: “Getting started is quick and easy, and our friendly team is always here to help you on your journey.”
- Easy to cite: “Setup takes about 20 minutes for a single site: create an account, add the tracking snippet to your site header and connect your analytics. Teams with several sites usually finish in an afternoon.”
The second version names the subject, gives a number, lists the steps and states the condition. A reader, or an answer engine, can lift it without rewriting it. The figures must be true for your product, of course. An invented number is worse than none.
Step 4: Make your facts easy to verify
An answer engine that shows its sources has every reason to prefer pages whose claims are specific and supported. Give it that reason.
- State numbers, prices, dates, limits and versions exactly, and say when they were last checked.
- Link to primary sources for claims you did not originate, such as standards, official documentation or published research.
- Publish first-party information only you can provide: your own data, tests you ran, documentation of how your product works, or clear policies.
- Keep the same facts consistent across your site. If your pricing page and your comparison page disagree, neither looks reliable.
- Show who wrote the page and why they are qualified to write it.
Step 5: Keep important pages fresh
Because Perplexity searches as it answers, recently updated pages can be picked up quickly. The flip side is that outdated pages can be cited with outdated facts, or passed over for a competitor’s newer page.
Put your most important pages on a refresh schedule. Update facts that change, such as pricing, features, integrations and availability, show a real update date, and let search engines know about changed URLs through your sitemap or IndexNow where supported. Do not change dates without changing content; that misleads readers and does not make a page more useful.
Step 6: Show up where Perplexity already looks
For many questions, the most-cited sources are not vendor sites. They are discussions, reviews, videos and articles by independent writers. If that is what you see for your questions, a citation on your own site is only part of the goal.
- Review and directory sites: claim your profiles on the platforms that appear in your answers and keep descriptions, pricing and categories accurate.
- Communities: where your audience asks questions, contribute useful answers under your real identity. Promotional posting tends to backfire and is often removed.
- Publications: offer expertise, data or commentary to the publications that already appear for your topic.
- Video: if videos are cited for your questions, a clear, well-titled explainer with an accurate description can become a source in its own right.
The guide on how AI search engines choose sources explains why claims echoed by several independent sites are safer for an engine to repeat. To gauge how your link profile compares with the domains Perplexity cites, use the free domain authority checker.
Step 7: Keep the technical basics tight
At the time of writing, Perplexity does not offer a webmaster console like Google Search Console, so you cannot submit pages to it directly. Make them easy to find and read instead, and use the AI search optimization checklist for a full audit.
- Serve the main content in the HTML, not only through client-side scripts.
- Keep pages fast and stable, with no errors for the bots you want to allow.
- Use descriptive titles and headings that say what the page answers.
- Keep your XML sitemap current and your canonical tags correct.
- Add structured data that matches the page, such as Article, Organization and FAQPage. The free FAQ schema generator handles the FAQ part.
- Consider an llms.txt file that lists your best pages. It is an emerging convention rather than a confirmed signal, but it is cheap to add with the free llms.txt generator.
Step 8: Measure citations, crawls and visits
Perplexity is easier to measure than most engines, because citations are visible and its visitors usually arrive with a perplexity.ai referrer. Combine three views:
- Citations: re-run your question list on a schedule and record whether you are cited, which page is cited and who else appears. Judge trends across weeks. The method in how to track brand visibility in ChatGPT works the same way here.
- Crawls: check your server logs for PerplexityBot and Perplexity-User and verify them against Perplexity’s published IP ranges. A page that is cited but rarely crawled, or crawled often but never cited, tells you where to look. The AI crawler analytics docs show how to read this data.
- Visits: segment referral traffic from Perplexity in your analytics and look at which pages it lands on and whether those visitors convert.
If a page that used to be cited drops out, follow the four-week process in how to win back lost citations.
Common mistakes
- Blocking all AI bots at the CDN, then wondering why Perplexity never cites you.
- Writing one long page for a whole topic instead of focused pages for the questions people ask.
- Leaving old prices or discontinued features live, which Perplexity can find and repeat.
- Posting promotional comments in communities instead of useful answers.
- Measuring from a single answer instead of repeated runs across many questions.
Frequently asked questions
How long does it take to get cited by Perplexity?
There is no published timeline. Because Perplexity searches the web as it answers, a new or updated page can be cited once it has been crawled and is judged relevant, which can happen quickly. Check your logs for PerplexityBot visits to the page, then re-run your questions over the following weeks.
Is PerplexityBot used to train AI models?
Perplexity says PerplexityBot is used to surface and link websites in its search results and is not used to crawl content for training AI foundation models. If you want to be cited, allowing it is the main access decision you need to make.
Do I need to rank in Google to be cited by Perplexity?
Not directly. Perplexity runs its own crawler and retrieval, and it has not said that Google rankings play a part in its choices. In practice, the qualities that help in Google, such as crawlable, relevant and trustworthy pages, also help in Perplexity, so solid SEO is a good foundation.
Can I pay to be cited by Perplexity?
There is no public way to pay for a citation in a Perplexity answer. Any advertising it runs is labeled separately from the cited sources, and this guide is about the organic citations inside the answer itself.
Why does Perplexity cite my competitor and not me?
Compare the cited page with yours. Common reasons are a more direct answer, more specific facts, a newer update, a closer match to the exact question, or a source type the engine favors for that topic, such as a review site. Fix access issues first, then close the content gap on substance.
Does Perplexity send real traffic?
Yes. Citations are clickable links, and visits usually arrive with a perplexity.ai referrer. Volume depends on your audience and topic, so measure it in your own analytics rather than relying on general estimates.
See your Perplexity citations with aeoora
aeoora asks your buyer questions in Perplexity every day, along with ChatGPT, Google AI Overviews, AI Mode, Gemini, Claude and Copilot, and records which domains and pages each answer cites, for you and for your competitors. It flags lost citations, measures visits from Perplexity with a cookie-free snippet, and on the Signal and Scale plans verifies PerplexityBot visits from your server logs. Plans are Shine at $29, Signal at $69 and Scale at $299 a month, with 25% off yearly and a 2-day free trial.
Start with the free AI visibility report to see how you show up for 25 buyer questions, or start a 2-day free trial and set up your prompts with the quickstart guide. New to the terms? The AI search glossary explains them.


