Perplexity SEO: how Perplexity finds and cites sources, and how to become one

Perplexity SEO is the work of getting your pages chosen as sources when Perplexity searches the web to answer a question. Start by allowing PerplexityBot, the crawler Perplexity uses to surface and link websites in its results, then make your pages the most specific and current answer to the questions your buyers ask.

Perplexity is a practical engine to optimize for because it shows its sources on every answer. Its help center says it searches "the internet in real-time" and that "each answer includes numbered citations linking to the original sources" (Perplexity Help Center). So you can see quickly whether you are cited, and who is cited instead.

How Perplexity builds an answer

From Perplexity's own description, the flow is simple. A question comes in, Perplexity searches the web, reads what it finds, writes an answer and attaches numbered citations to its sources. Follow-up questions keep the context of earlier ones. Pro Search gives longer answers and lets the user pick which AI model writes them, and Deep Research "performs dozens of searches automatically, reads hundreds of sources" before it writes a report (Perplexity Help Center).

We draw two practical conclusions from that, which Perplexity does not state itself. First, one user question can turn into many searches, so a page that answers a narrow sub-question can still be cited. Second, because the answer is written from a live search, a page with current facts has an edge over an older page that says the same thing less precisely.

PerplexityBot and Perplexity-User are different bots

Perplexity documents two user agents in its crawler guide, and they behave differently:

PerplexityBot is the one that decides whether your pages are in Perplexity's search results. Perplexity-User acts more like a person's browser. A robots.txt rule against it does little, and blocking it at the firewall stops Perplexity from fetching your page in the middle of answering someone, which is one of the ways a page ends up linked.

Set up access in robots.txt and your firewall

A minimal robots.txt rule that lets Perplexity crawl the whole site looks like this:

User-agent: PerplexityBot
Allow: /

If you have a broader rule, such as User-agent: * with Disallow lines for some folders, check that it does not cover the pages you want cited. Perplexity says changes can take up to 24 hours to show in its systems (Perplexity).

robots.txt is only half the job. Perplexity notes that sites behind a web application firewall "may need to explicitly whitelist" its bots, and recommends combining user agent matching with IP verification against the published ranges rather than trusting the user agent string alone (Perplexity). If your CDN or firewall has a setting that blocks AI crawlers, check whether it covers PerplexityBot.

Our free AI crawler checker reads your robots.txt for PerplexityBot, Perplexity-User and 15 other crawlers. It cannot see firewall rules, so also look for PerplexityBot requests in your server or CDN logs.

What gets a page cited in Perplexity

Perplexity does not publish how it ranks sources. These are the levers you can support with evidence.

Be specific enough to sit next to a citation number

Perplexity attaches numbered citations to statements in its answer. A page that backs one clear statement with a number, a date or a named source gives it something to attach a citation to. The GEO paper by Aggarwal et al. (2023) found that adding quotations, statistics and cited sources raised visibility in generated answers by up to about 40% on the authors' benchmark, while keyword stuffing did little. That benchmark was not Perplexity itself, so treat it as direction rather than a forecast.

Keep it current and dated

Answers are written from a live search, and AI citations churn: Profound puts the half-life of an AI citation at about 11 days (Profound). Put a visible updated date on pages with prices, versions or rankings, keep those facts correct, and revisit your most important pages when something changes.

Show up where Perplexity looks

In Profound's analysis of 680 million citations from August 2024 to June 2025, Reddit was the most-cited single source for Perplexity, at 6.6% of its citations. The lesson is not to flood threads with links. It is that genuine discussions which name your product, along with the roundups and comparison pages Perplexity cites for your category, feed its answers. Answer questions in communities under your own name and say who you work for, and send roundup authors your current facts.

Cover the sub-questions

Deep Research runs dozens of searches for one request (Perplexity Help Center), so a single buyer question becomes lookups on pricing, setup, integrations and alternatives. List the sub-questions behind each important query and make sure one of your pages answers each in its first sentences, under a heading that matches it.

Common Perplexity SEO mistakes

• Blocking PerplexityBot by accident, through a blanket rule or a CDN bot setting.
• Trying to block Perplexity-User in robots.txt, which Perplexity says it generally ignores.
• Allow-listing by user agent string only, without checking the published IP ranges.
• Testing with your brand name. Asking Perplexity "what is [your brand]" will describe you and says nothing about category questions.
• Assuming a ChatGPT citation means a Perplexity one. Profound found that how long a page stays cited on one engine says almost nothing about how long it stays cited on another, with correlations of 0.03 to 0.09; in its words, "The engines decide independently."

How to measure Perplexity visibility

1. List 10 to 20 questions your buyers ask, without your brand name in them.
2. Ask each one in Perplexity and save the numbered sources. Repeat a few days later and keep the domains that recur.
3. Record mention and citation separately. Perplexity can name you in the text and cite someone else's page.
4. Watch referral visits from perplexity.ai in your analytics.

XeoRank can run steps 2 and 4 for you. When you start a scan, it asks Perplexity your questions and saves the sources it cites, alongside ChatGPT and Gemini (grounded in Google Search), and shows each engine on its own row, next to an overall citation rate. Starter is $19 a month and Pro is $49 a month; see pricing for what each plan includes. The GA4 report, on every plan, shows AI referral traffic from Perplexity, ChatGPT, Claude, Gemini and Copilot against the previous period.

What it does not do: XeoRank asks Perplexity through Perplexity's API, so its answers can differ from what a signed-in user sees in the app. It does not track Deep Research reports, cannot see inside Perplexity's ranking, and cannot make Perplexity cite you.

It shows where you stand and which pages to fix.

Where to start

Check robots.txt and your firewall for PerplexityBot today, ask Perplexity your five most important buyer questions, and list the cited domains. For the steps that apply to every engine, read how to get cited by AI. For ChatGPT, read how to rank on ChatGPT.