Perplexity SEO: How to Get Your Site Cited as a Source

Blog · Perplexity SEO

Perplexity SEO means setting up your website so that Perplexity can find it, understand its content and link to it as a numbered source in its answers. According to its help centre, Perplexity searches the web in real time for every question, summarises what it finds and backs the answer with citations that lead straight to the original page. Underneath sits Perplexity's own search index, which is filled by a crawler called PerplexityBot. Perplexity says this crawler respects robots.txt and is not used to train AI models. To be considered as a source, your site has to be reachable for PerplexityBot and offer content that can be quoted section by section. How Perplexity chooses between several suitable pages is something the company does not disclose. Serious work on Perplexity SEO therefore starts with what is documented, and with your own measurements.

In brief

Perplexity searches the live web for every question and backs its answer with numbered, linked sources.
PerplexityBot fills the search index, respects robots.txt according to Perplexity and is not used for AI training.
Pages are broken into passages: individual sections are retrieved and ranked, not just whole documents.
Perplexity publishes no list of ranking factors, so you measure success through repeated queries and server logs.

Published 25 Sep 2026Nikolai Schöbel and Jeremias Burger8 min read

Nikolai Schöbel and Jeremias Burger

Co-founders of Scalableloops GmbH. Nikolai Schöbel leads online marketing and AI strategy, Jeremias Burger the AI architecture. Both build AI systems and train teams on them in their own agency work.

On this page

What is Perplexity AI and how does it build an answer?
Where does Perplexity get its sources from?
Which crawlers does Perplexity send to your website?
Should you block PerplexityBot in robots.txt?
How does Perplexity SEO work in practice?
What does Perplexity not reveal about Perplexity SEO?
How do you measure the success of Perplexity SEO?
Frequently asked questions
How to get started with Perplexity SEO
Where the information on this page comes from

Basics

What is Perplexity AI and how does it build an answer?

Perplexity AI is an answer engine: you ask a question in full sentences and get a summarised answer instead of a list of links. Its help centre describes the process in three steps. A language model first interprets the question. Perplexity then searches the internet in real time, gathering information from articles, websites and journals. Finally, it condenses the most relevant points into an answer that is easy to read.

One feature matters most for businesses: every answer includes numbered citations that link to the original sources. Perplexity presents this as transparency, so that users can verify statements or read further. Unlike a classic results page, users can see exactly which pages a particular statement relies on. If your page is among them, you are cited in plain sight. If it is not, you are absent from the answer, however well you rank elsewhere.

This visible attribution is the heart of Perplexity SEO. The goal is not a position in a list but being the evidence behind a statement. For the same question applied to ChatGPT, see our article on ChatGPT SEO.

Index

Where does Perplexity get its sources from?

Perplexity draws on its own search index. In September 2025, when it launched a search API for developers, the company published a research article describing how its search works. By its own account, the API runs on the same infrastructure as the public answer engine, and the index covers hundreds of billions of web pages.

Two points from that article are especially useful in practice. First, Perplexity works below the level of whole documents. Pages are split into self-contained passages that are retrieved and scored individually for each query. The aim is to hand the language model only the passage that answers the question, without the rest of the page. A paragraph that answers a question fully on its own is therefore better placed than an answer scattered across several parts of a page.

Second, Perplexity cares about freshness. The index is updated continuously, with tens of thousands of indexing operations per second according to Perplexity. A machine learning model decides when each URL should be fetched again, based on how important it is and how often it is likely to change. For you, this means that pages which are maintained regularly and show a clear update date fit this approach better than stale content.

Hundreds of billions

of web pages are covered by Perplexity's own search index, which powers both its answers and its search API.

Perplexity, 25 Sep 2025; InfoQ, 30 Sep 2025

Crawlers

Which crawlers does Perplexity send to your website?

Perplexity documents two user agents that you may see in your server logs. They serve different purposes and treat robots.txt differently. The table reflects Perplexity's documentation as of 25 September 2026.

Aspect	PerplexityBot	Perplexity-User
Purpose according to Perplexity	Surfacing and linking websites in Perplexity's search results	Supporting actions a user takes in Perplexity, such as fetching a specific page
AI training	Not used to collect content for AI foundation models	Not intended for web crawling or AI training
robots.txt	Respected	Generally not followed, because a user requested the fetch
Identification	PerplexityBot in the user agent string, official list of IP addresses	Perplexity-User in the user agent string, separate list of IP addresses
What it means for you	Decides whether your pages enter the index and can be cited	Fetches pages directly when someone asks about them in Perplexity

Access

Should you block PerplexityBot in robots.txt?

Not if you want to appear as a source in Perplexity. PerplexityBot is the search crawler, not a training crawler. Blocking it removes your pages from the pool Perplexity draws its sources from. Unlike OpenAI, Perplexity does not document a separate training user agent that you would need to manage on its own. For an overview of other providers' crawlers, see our article on AI SEO.

Look beyond robots.txt, too. Perplexity explicitly advises site owners to check both the user agent and the IP address in their firewall rules, and publishes lists for this purpose. There is a practical reason: many security services now block AI crawlers on their own. Since 1 July 2025, Cloudflare has asked every newly set up domain whether AI crawlers should be allowed, and offers existing customers a one-click block. Whether PerplexityBot is caught by your configuration can only be seen in your provider's settings or in your logs.

There is also a dispute you should know about. On 4 August 2025, Cloudflare accused Perplexity of continuing to fetch blocked sites with undeclared crawlers and rotating IP addresses, and removed it from its verified bots programme. Perplexity publicly rejected this, saying the traffic consisted of fetches triggered by users and that Cloudflare had attributed some unrelated traffic to Perplexity. The dispute is unresolved. The takeaway for your decision: robots.txt controls whether you are in the index, but it is not a technical barrier. If you genuinely need to protect content, you need rules at firewall level.

Actions

How does Perplexity SEO work in practice?

Perplexity's documentation and its technical article point to measures that rest on described behaviour rather than guesses about secret factors. The order is deliberate: without access and readable content, nothing else takes effect.

01
01
Secure access
Check robots.txt, your firewall and your host's bot protection for PerplexityBot. Ask for confirmation that requests from the official IP ranges are not blocked.
02
02
Core content in the HTML
Services, product data, contact details and key messages belong in the HTML your server delivers. Perplexity does not document whether PerplexityBot executes content loaded later by scripts. Only what sits directly in the source code is safe.
03
03
Sections that stand alone
Because Perplexity splits pages into passages and scores them separately, each section should answer one question completely: the question as a subheading, the answer in the first sentence, details afterwards.
04
04
Show that content is current
Maintain important pages regularly and display the date of the last update. Perplexity schedules refetches according to a URL's importance and expected rate of change.
05
05
Evidence over claims
Specific figures with a source, unambiguous product details and tables can be cited as evidence. Generic marketing statements give an answer engine nothing to build a statement on.
06
06
Presence beyond your own site
Perplexity draws sources from the whole web, not only from manufacturers' pages. Trade articles, industry directories and press coverage that describe your company consistently increase the number of places where you can show up as evidence.

Limits

What does Perplexity not reveal about Perplexity SEO?

Perplexity explains how its search is built technically, but not which criteria decide between several suitable pages or where a page is cited. Neither the crawler documentation nor the help centre contains a list of ranking factors, a statement on backlinks or domain strength, or a process for site owners to request inclusion.

Be careful, then, with guides that promise precise weightings. Such claims rest on observations by individual vendors, not on statements from Perplexity. Three things are solid: PerplexityBot must be allowed to fetch your page, your content must be quotable as a self-contained passage, and it should be current. Everything else is best tested against real answers.

Perplexity's crawler documentation also makes no mention of llms.txt files, structured data or special AI text files. What llms.txt can do in general is covered in our article on llms.txt. For how Google officially describes its AI summaries, see our article on Google AI Overviews.

Measurement

How do you measure the success of Perplexity SEO?

Because Perplexity attaches sources to every answer, visibility there is easier to check than in many other AI systems. Collect the questions your customers ask before they get in touch, from general research to comparing providers, and put them to Perplexity. Note whether your brand is named in the text, whether one of your pages appears among the sources, and which competitor or portal pages show up instead.

Answers vary from one query to the next. Repeat the same questions several times and over several weeks before drawing conclusions. Your server logs add the other half of the picture: whether PerplexityBot fetches your pages at all, and which ones. Match the IP addresses against the official list, because a user agent string can be faked.

The source list also tells you which third-party pages Perplexity relies on for your topic. That is a list of places where your company ought to appear. How to measure AI visibility across several systems is described in our article on measuring AI visibility. For a full analysis with a competitor comparison, talk to our GEO agency.

Frequently asked questions

What is Perplexity SEO?

Perplexity SEO covers everything that makes your website reachable, understandable and citable for Perplexity. The goal is to be linked as a numbered source in Perplexity's answers. It is one part of Generative Engine Optimization.

How does Perplexity choose its sources?

For each question Perplexity searches its own index, splits pages into individual passages and scores them in several stages. The exact selection criteria are not published. What is documented is access via PerplexityBot and the weight Perplexity places on current content.

Does Perplexity use my content to train AI models?

According to Perplexity, PerplexityBot is not used to collect content for AI foundation models. Its purpose is to surface and link websites in Perplexity's search results.

What happens if I block PerplexityBot in robots.txt?

PerplexityBot respects robots.txt, so your pages drop out of Perplexity's search index and can no longer be cited. Fetches that a user explicitly requests run under the Perplexity-User agent, which according to Perplexity generally does not follow robots.txt.

How can I tell whether Perplexity visits my website?

Your server logs will show the user agents PerplexityBot and Perplexity-User. As a user agent can be faked, check the IP address against the lists Perplexity publishes for this purpose.

Do I need separate optimisation for Perplexity on top of SEO?

The foundations overlap: reachable, well structured and up to date pages. On top of that come allowing PerplexityBot, sections that answer a question on their own, and measuring through real answers instead of rankings.

Next steps

How to get started with Perplexity SEO

01
Check access
Open your robots.txt and ask your host whether a firewall or bot protection blocks PerplexityBot.
02
Review your logs
Find out which pages PerplexityBot has fetched in recent weeks and match the IP addresses against the official list.
03
Test real questions
Put ten typical customer questions to Perplexity and record which sources are cited and whether your pages are among them.
04
Rework your key pages
Phrase subheadings as questions, put the answer at the start of each section and show the date of the last update.

Perplexity shows openly which pages it relies on. That makes visibility there testable: you can see whether your company appears, and you can see who is cited in your place.