Perplexity SEO means making your pages easy for Perplexity to find, read, and cite. Perplexity answers questions with numbered links to the sources it used. When one of those links is yours, a customer can click straight through.
This guide sticks to what Perplexity documents about its crawlers, answers, and source labels. It shows how to check that Perplexity can reach your site. It also covers how to write pages it can quote, with a worked example.
Reporting note: We checked this guide against Perplexity's public docs on November 21, 2026.
Key takeaways
- Perplexity cites its sources with numbered links in each answer
- PerplexityBot follows robots.txt, but Perplexity-User generally doesn't
- A plugin or CDN setting can block Perplexity without you knowing
- Specific, current facts are easier to quote than sales copy
- Judge progress with referrals and a monthly question log, not one answer
How does Perplexity find and cite pages?
Perplexity's help center describes four steps behind each answer. It reads the question, searches the web, sums up the results, and cites sources.
Each answer includes numbered citations that link to the original pages. For harder questions, Pro Search can run several searches and combine the results. Perplexity says Pro Search now runs on its own when your plan includes it.
That gives a business owner three jobs:
- Let Perplexity's crawlers reach your pages
- Answer real customer questions in plain text
- Give facts that are specific, current, and easy to check
Perplexity doesn't publish a list of ranking factors. So treat any tip that claims to know the formula with care. This guide separates what Perplexity documents from our own advice.
Perplexity's two crawlers, explained
Perplexity's crawler documentation names two user agents. Each one has a different job, and each follows different rules.
Scroll sideways to read the full table.
| Crawler | What it does | Follows robots.txt? | Used to train AI models? |
|---|---|---|---|
| PerplexityBot | Finds and links websites in Perplexity's search results | Yes, per Perplexity's help center | No, per Perplexity's docs |
| Perplexity-User | Visits a page when a person's question needs it | Generally no, since a person asked | No, per Perplexity's docs |
Perplexity says PerplexityBot isn't used to crawl content for AI foundation models. Its job is search. So blocking it mostly limits how your pages show up in Perplexity's answers.
Perplexity-User works differently. It fetches a page because someone asked a question that needs it. Perplexity says it generally ignores robots.txt for those requests, since a person made them.
Both crawlers have published IP ranges, listed in JSON files on Perplexity's site. Perplexity also says robots.txt changes can take up to 24 hours to apply.
In practice, you control PerplexityBot with robots.txt. You control Perplexity-User mainly through your firewall. If you want Perplexity to cite you, allow both.
What happens to a site that blocks PerplexityBot? Perplexity's robots.txt article says it won't index that site's full or partial text. It may still index the domain, the headline, and a brief factual summary.
Can Perplexity reach your website?
A site can block AI crawlers without the owner knowing. A security plugin or a CDN setting can do it quietly. Run these checks in order.
- Open your robots.txt file and look for PerplexityBot or a blanket Disallow rule
- Check your CDN, host, or security plugin for a setting that blocks AI bots
- If you use a firewall rule, match the user agent and Perplexity's IP ranges together
- Ask your host to search the server logs for PerplexityBot visits
- Wait at least a day after any change, then check the logs again
Your robots.txt file lives at the root of your domain. Type your domain followed by /robots.txt into a browser to see it. If it says Disallow: / under User-agent: *, it blocks every crawler that follows the rules.
In your server logs, each visit shows a user agent string. PerplexityBot's string includes PerplexityBot/1.0, and Perplexity-User's includes Perplexity-User/1.0. Perplexity notes the version number may change, so search for the name alone.
Perplexity's docs give firewall steps for Cloudflare and AWS WAF. They warn that custom rules can't skip Cloudflare's Bot Fight Mode. So it can still challenge Perplexity's crawlers even after you allow them.
Want the full robots.txt walkthrough for other AI crawlers too? Our guide to AI crawlers and robots.txt covers the other major bots.
Blocking PerplexityBot can be a fair business choice. Just make that choice on purpose, not by accident.
What makes a page easy for Perplexity to cite?
Perplexity hadn't published a content guide for site owners when we checked. So the tips below are our advice, based on how its answers work. They also match Google's guidance on helpful, people-first content.
Perplexity sums up pages and links to them. A page is easier to quote when its answer is plain, specific, and checkable. Here's what that looks like on a service business site:
- A short, direct answer under each question heading
- The facts customers ask about, like service area, hours, and real price ranges
- A named author and a visible last-updated date
- Links to the sources behind any claim or figure
- Key facts in the page text, not only in images or PDFs
- The same business details on every page and profile
Specific beats general. "We repair tankless water heaters in Brandon and Riverview" is easy to quote. "Quality service at fair prices" gives an answer engine nothing to use.
Use headings that match how people ask. "How much does a water heater swap cost?" works better than "Pricing." Then answer it in the first sentence below the heading.
Write each answer so it still makes sense on its own. Perplexity may quote one section, not the whole page. A section that starts with "As mentioned above" loses its meaning out of context.
Many of the same steps help with ChatGPT too. Our guide to getting cited in ChatGPT search covers that side.
What do Perplexity's source labels mean?
Some cited sites show a shield icon with a label. Perplexity's help center lists three labels: Government, Academic, and Trusted. Each label applies to a whole website, not to single pages.
Perplexity says a source review process sets these labels. It looks at whether a site corrects its mistakes and names its authors. It also checks whether a site keeps news apart from ads and opinion.
Perplexity says partnerships and payments don't affect a site's label. It calls unlabeled sites "not yet rated," not low quality. Most small business sites won't carry a label, and that's normal.
The review points are still worth copying. Put a real name on each guide and note corrections when you fix an error. Keep sales copy apart from advice.
Selling products? Read the Instant Buy rules
Perplexity's Instant Buy lets US users buy from some merchants inside Perplexity. Orders go straight to the merchant. It only works with merchants that fit Perplexity's checkout system.
Perplexity says advertisers can't pay to list products in its related products section. It also says it's more likely to recommend merchants with deeper product details. It names availability, reviews, pricing, and specifications.
For a store, that means complete product pages and accurate stock. For a service business, it's a nudge to publish clear service and price details too.
A worked example: a plumber checks Perplexity
Here's a hypothetical example. Say a plumbing company in a mid-sized city wants to show up in Perplexity's answers. The owner has one free afternoon and follows these steps.
- Write five questions customers really ask, like the cost of a water heater swap
- Ask each one in Perplexity and note which sites it cites
- Check robots.txt and the firewall, and find a plugin that blocks AI bots
- Turn off that block, then confirm PerplexityBot visits in the logs a day later
- Rewrite the water heater page with a short answer, price range, and service area
- Add the author's name, an updated date, and links to local permit rules
- Ask the same five questions next month and log what changed
In this example, most cited sites were directories and a few rival service pages. Those pages answered the question in the first lines. The plumber's own page buried the answer under a sales pitch.
After the rewrite, the answer sits right under the heading. The price range shows what drives the cost up or down. The service area lists the towns by name.
The owner also found an old directory listing with last year's hours. They updated it, since cited directories can repeat stale facts.
No step here can force a citation. Each one removes a reason Perplexity might pass the page over.
How can you tell if Perplexity sends you visitors?
Clicks on Perplexity's citations show up as referral traffic in Google Analytics 4. Look for perplexity.ai as the source. Our guide to tracking AI search traffic in GA4 shows how to group these visits.
Server logs add a second view. A Perplexity-User visit means someone's question pulled in your page. That can happen even when the person never clicks.
Keep a simple monthly log in a spreadsheet. Answers can change from day to day, so watch the trend, not one result. These columns are enough:
Scroll sideways to read the full table.
| Column | What to write |
|---|---|
| Date | The day you asked the questions |
| Question | The exact words a customer would use |
| Sites cited | The domains in Perplexity's numbered citations |
| Your page cited? | Yes or no, plus which page |
| Wrong facts | Any old price, hours, or service area in the answer |
| Fix made | What you changed and where |
The wrong facts column matters most. If Perplexity repeats an old price or a closed location, find where that fact lives. Fix it on your site and on any profile that still shows it.
Perplexity SEO mistakes to avoid
Most problems come from a handful of habits. Watch for these:
- Blocking PerplexityBot by accident through a plugin or CDN setting
- Allowing the user agent but not the IP ranges in a firewall rule
- Paying anyone who claims they can sell you a Perplexity citation
- Publishing thin, AI-written pages that repeat what every other site says
- Letting hours, prices, or service areas differ across your site and profiles
- Judging results on one answer instead of a monthly trend
Recheck access after any big site change. A new theme, a plugin update, or a host move can reset firewall settings. A five-minute look at robots.txt and your logs catches it.
Perplexity says payments don't affect its source labels or its related product listings. Be wary of anyone selling a shortcut.
Also skip "ranking factor" lists for Perplexity. Perplexity hasn't published one, so those lists are guesses. Spend the time on clearer pages instead.
