How to Get Cited in Perplexity: Sources, Citations and Optimization

To show up in Perplexity, one of the fronts of GEO, allow PerplexityBot in robots.txt and in your firewall, publish pages that answer specific questions clearly and keep them up to date. Perplexity itself says PerplexityBot exists to show and link sites in its results. How it chooses which of the available pages to cite is not public, and a citation is never guaranteed.

Quick summary

• Perplexity searches the web to answer and shows the sources next to the answer. Becoming an eligible source is the first goal.
• Perplexity documents two agents: PerplexityBot (search, respects robots.txt) and Perplexity-User (user-initiated access, which generally ignores robots.txt).
• According to the company, PerplexityBot is not used to collect content for training foundation models.
• Studies show little overlap between the sources cited by different platforms: optimizing for one does not guarantee the others.
• Measure with repeated manual tests and with session sources in GA4, because Perplexity is not part of Google Analytics' default "AI Assistant" channel.

If you have not read the overview yet, start with SEO, AEO and GEO: what is the difference. The beginner version of this same topic is how to show up in ChatGPT.

How to read this post: if you are not technical, read the sections "How does Perplexity work in practice?" and "How do I allow access, step by step?" and use the checklist. If you are an SEO professional, go to the agent table, the log check and the measurement topic.

How does Perplexity work in practice?

Perplexity is an "answer engine": you ask a question, it searches the web for pages, uses passages from them to write an answer and shows the sources alongside the text, in numbered notes that lead to the original site. Unlike ChatGPT, where search is one of two paths, Perplexity's format is built around search with citation.

In plain English: imagine a student who, for every question, goes to the library, pulls a few books and writes a summary noting which book each sentence came from. Your site only makes it into the summary if it is on the shelf (reachable by the bot) and if the right passage is the most useful one for that question.

What the company does not publish: the formula for selecting and ordering sources. In its developer material, Perplexity shows that its API's search results carry date metadata (date and last_updated), according to the API filters documentation. That suggests the product treats a content's date as relevant information, but it does not prove how much weight it carries in the choice.

Which Perplexity bots exist and what do they do?

There are two, and the difference defines what you can block or allow. The official Perplexity crawlers page describes:

Agent	What it is for (according to Perplexity)	Respects robots.txt?	Used to train models?
PerplexityBot	Showing and linking sites in Perplexity search results	Yes; the company advises allowing it so the site appears in results	No, according to the company
Perplexity-User	User actions: when answering, it may visit a page to give an accurate answer	Generally ignores it, since a person requested the access	No, according to the company

Both have IP ranges published as JSON, and Perplexity provides setup instructions for Cloudflare and AWS WAF that combine user agent and IP.

In plain English: robots.txt is a sign at the building's front door saying who may enter. PerplexityBot reads the sign and obeys. Perplexity-User is a visitor who comes in because a person invited them, so the sign at the door does not apply in the same way.

For SEO professionals: because Perplexity-User generally ignores robots.txt, blocking PerplexityBot does not stop a page from being read when someone requests it inside the product. If you need to restrict that access, the barrier has to be on the server or in the WAF (for example, authentication), not in robots.txt. And note: a user agent alone can be spoofed, which is why the WAF setups Perplexity describes combine user agent and IP.

How do I allow access, step by step?

Allowing access is the part you fully control. Do it in this order:

1. Check your robots.txt (yoursite.com/robots.txt). Look for Disallow: / under User-agent: * and for specific rules for PerplexityBot.
2. Explicitly allow PerplexityBot:

   User-agent: PerplexityBot
   Allow: /
   

3. Review the WAF/CDN. "Bot fight" rules, anti-scraping protection and geoblocking often block AI bots. Create an exception that combines the user agent and the IP from the official list.
4. Search the logs. If you have access to the server log, filter for PerplexityBot and look at the response codes:

   grep -i "PerplexityBot" access.log | awk '{print $9}' | sort | uniq -c
   

   Lots of 403 or 429 responses indicate blocking. A 200 indicates the bot reads your page (the status field varies with your log format).
5. Make sure the HTML is readable. The main content should be in the HTML delivered by the server, without depending on a login, pop-up or late JavaScript loading. Confirm with the browser's "view page source," not "inspect."
6. Keep your sitemap and internal links up to date. It is the common discovery path for any search engine. See the technical SEO basics.

What kind of page is most likely to be cited?

There is no official list of "preferred types." What can be said honestly is that Perplexity answers specific questions, and pages that answer a specific question in an extractable way are better candidates than generic pages. That is an inference from how the product works, not a documented rule.

Good practices that make sense for any answer system with citations:

• One question, one direct answer right below the heading, with the details after.
• Data with a source and a date, because the reader (and the system) needs to verify. See how to create original research with your own data.
• A visible update date on topics that change. Perplexity exposes dates in its API results, which is a good reason to keep them correct.
• Tables and lists that allow a passage to be clipped without losing its meaning.
• Clear authorship and sources, the basis of E-E-A-T and trust for AI.

The writing how-to is in how to structure content that LLMs can cite. To understand why a specific passage gets chosen, read how LLMs search: RAG and grounding.

For SEO professionals: Ahrefs analyzed cited sources across ChatGPT, Perplexity and Google AI Overviews and reported that only 7 of the 50 most-cited domains appeared on all three platforms, which means about 86% were exclusive to one of them, according to Ahrefs' evidence review (September 2025). Practical takeaway: measuring one platform does not let you infer the others.

How do I know if Perplexity is citing me and sending visits?

Measure in two ways: by asking Perplexity itself and by looking at traffic. Neither is perfect, which is why you use them together.

1. Repeated manual test. Pick 10 to 20 questions your customers would ask, ask them in Perplexity, and for each answer note whether your brand was cited, whether your domain appeared in the sources and which domains appeared instead. Repeat on different days, because sources change. The full method is in how to find out if AI recommends your brand.

2. GA4 traffic. GA4's default "AI Assistant" channel lists ChatGPT, Gemini, Deepseek, Copilot and Grok in the official definition, and does not mention Perplexity, according to the channel groups documentation. So create a custom channel group:

1. In Admin > Channel groups, create a new group (starting from a copy of the default one).
2. Add a channel "AI - Perplexity" with the condition Source matches regex: perplexity.
3. Move that channel above "Referral". GA4 assigns traffic to the first channel whose rule it meets, according to the custom channel groups documentation.

Two caveats: the label depends on the browser or app sending the visit's source (without it, the session falls into Direct), and some of the people who see your brand in Perplexity arrive later through a search for your name, with no AI label. The full reading of indicators is in how to measure AI visibility.

Hypothetical example: a SaaS that showed up elsewhere but not in Perplexity

Hypothetical example: "Shelfwise," a fictional inventory control tool for small shops, has a guide on "how to calculate reorder point." In Perplexity, the question "how to calculate reorder point for a small shop" cites three competitors, never Shelfwise.

The team follows the checklist. There is no block in robots.txt. In the log, however, PerplexityBot gets 403 since the host turned on an anti-bot protection. They create the exception by IP and user agent. Then they rewrite the guide: it opens with the formula and a numeric example, and adds a table with three scenarios and the date of the last review.

Two weeks later, the log shows 200, and Shelfwise repeats the test 10 times over a week. If it shows up in 2 out of 10, that is not a result: the sample is too small to claim more than "it appeared a few times." The point of the example is the order: first access, then content, then repeated measurement.

Frequently asked questions

Do I need to block Perplexity-User to protect my content?

robots.txt is not the right tool. Perplexity says this agent, because it acts at a person's request, generally ignores the file's rules. If the content is restricted, protect it with a login or server rules. If it is public and you want to appear, there is no reason to block access.

Does Perplexity use my content to train models?

According to Perplexity's documentation, PerplexityBot is not used to crawl content for foundation models, and Perplexity-User is not used to collect content for training. That is the company's statement, not an independent audit. If this is critical to you, read the official page and review it periodically.

Is there a way to submit my site to Perplexity?

I found no submission form or tool in the official crawler documentation. The documented path is allowing PerplexityBot in robots.txt and in the firewall. Page discovery follows the standard web: links, sitemap and mentions.

Does ranking first on Google help in Perplexity?

It can help indirectly, because good pages tend to have links and mentions, but there is no proof of a direct relationship and Perplexity does not detail its ranking. The Ahrefs data on low source overlap between platforms suggests each system chooses in its own way.

How often should I test?

For simple tracking, once a week or every two weeks, with the same questions and the same log. Run each question more than once, because answers vary. Small changes between one measurement and the next can be noise, not the result of your work.

Next step

If your focus was ChatGPT, go back to how to show up in ChatGPT. To turn testing into a routine with reliable metrics, move on to how to measure AI visibility: metrics, share of voice and attribution. Anyone who wants automatic tracking can use Contentor's AI Visibility, which checks, for the questions you choose, whether ChatGPT cites the brand. It monitors ChatGPT only, not Perplexity.

Read also:

• How to structure content that LLMs can cite
• robots.txt and AI crawlers
• How LLMs search: RAG and grounding