How to Rank in Perplexity and Get Your Brand Mentioned
Perplexity has no ranked list, so the steps people recommend need sorting: what Perplexity documents, what a study tested, what rests on correlation, and what has no support.
Angelique Swain · · 6 min read

On this page
Your sales team mentions that a prospect found a competitor through Perplexity, and by Friday the question has landed on your desk: how to rank in Perplexity? You try it yourself, search your own category and read the answer. Your brand isn't in it.
Here's what you're up against. Perplexity doesn't publish a ranked list. It retrieves pages, cites some of them and names a few brands in the answer, and each of those is a different result with a different fix.
Was your page never retrieved? Cited for one claim but never named? Let's sort the steps people recommend by what Perplexity documents, what a study tested and what rests on correlation alone.
How does Perplexity build an answer?
Perplexity interprets the question, searches the web and compiles the most relevant insights into an answer with numbered citations linking to the sources, according to its help center.
That process has no rank position to hold. A page can be retrieved, cited for a single claim, or passed over while the brand still gets named from another source. Firemark's guide to named, linked and cited separates the three.
Take a buyer who asks Perplexity for payroll tools for a ten person company. The answer might cite your pricing page in note [3] while naming only two competitors in the text. Your page was retrieved and cited, and your brand wasn't named. A report that says "we rank" can't describe that, which is why the counts below keep the outcomes apart.
Can Perplexity's crawler reach your site?
PerplexityBot is the crawler that surfaces and links sites in Perplexity search results. Perplexity recommends allowing it in robots.txt and permitting requests from its published IP ranges, per its crawler documentation.
Perplexity-User is a separate fetcher that visits a page when a user's question calls for it. Perplexity says it generally ignores robots.txt because a user triggered the request. Perplexity also says it doesn't use either bot to collect content for training foundation models.
If your pages never turn up in Perplexity answers, check access before anything else:
- Open robots.txt and look for a rule that disallows PerplexityBot, such as a
User-agent: PerplexityBotline followed byDisallow: /. - Check firewall and CDN rules against the IP lists Perplexity publishes for each bot, and match the user agent and the IP together, as Perplexity advises.
- Wait before you test again, because Perplexity says its systems can take up to 24 hours to reflect changes.
- Ask your host whether bots are filtered at the edge, because Perplexity notes that firewall changes can take time to propagate.
Access is the one lever Perplexity documents. An open door doesn't tell Perplexity to cite you. A block on PerplexityBot removes the crawler Perplexity says it uses for search results, though Perplexity-User fetches can still happen.
Do Google rankings carry over to Perplexity?
Perplexity's cited URLs overlapped heavily with Google's top 10 on short head terms: 65.07% of them appeared there, against 10.00% for ChatGPT, across 3,311 search terms in an Ahrefs study published in September 2025.

At the domain level, Ahrefs found 80.58% against 31.80%. The figures cover short-tail terms only, they describe overlap and not cause, and Ahrefs sells a competing tool. Read them as a sign that ordinary SEO may carry weight in Perplexity, not as a rule.
The study is one slice. Ahrefs describes it as part of a series, and in the earlier long-tail study Perplexity's overlap with Google's top 10 was 28.6%. Ahrefs notes the figures come from different studies and datasets, and the short-tail study was published over a year ago.
If a page already sits in Google's top 10 for a buyer question, that's the cheapest page to check in Perplexity first.
How to rank in Perplexity: what works?
Of the steps people recommend for Perplexity, only crawler access is documented by Perplexity itself. A few content changes have a lab study behind them, and most of the rest depend on correlation or assertion.
| Lever | What supports it | Strength |
|---|---|---|
| Crawler access | Perplexity's crawler documentation | Documented |
| Ranking well in Google | Ahrefs, published September 2025: 65.07% of Perplexity's cited URLs were in Google's top 10 (short-tail terms only) | Correlational |
| Adding quotations or statistics to source text | GEO paper, 200 example test on Perplexity.ai with supplied files | Tested in a lab setting |
| Keyword stuffing | Same paper: about 10% worse than baseline on word count, but higher on subjective impression | Tested in a lab setting, mixed |
| Placement on third-party lists | Asserted in agency guides, no study found | Not supported |
| Fresh content | No study found that isolates Perplexity | Not supported |
The only controlled test of content changes on Perplexity that turned up is the GEO paper, first posted in 2023 and published at KDD 2024. On a subset of 200 examples, the paper reports that adding quotations gave the largest gain on its position-adjusted word count, 22% over baseline. The measure tracks how much of an answer draws on a source and how early. Adding statistics gave the largest gain on its subjective impression score, 37%.

The test's scope is narrow. Perplexity didn't let a user specify source URLs at the time, so the authors uploaded source text as files and had Perplexity answer only from those files. The result shows how supplied text is weighted inside an answer, not whether Perplexity finds or cites a live page. To check any lever on your own pages, track a fixed set of questions, as Firemark's template for building a prompt set shows.
The two "not supported" rows aren't claims that the tactics fail. Appearing on a roundup could plausibly get a page retrieved more often, and fresh pages may get picked up sooner. Agency guides assert both, and no study read for these rows measured either one on Perplexity. Until one does, they're hypotheses to test on your own prompts.
What does the evidence not support?
No evidence supports a guarantee that any step gets a brand mentioned in Perplexity, because the same prompt returns different brand lists from run to run.
SparkToro's study of AI recommendations, published in January 2026, had 600 volunteers run 12 prompts through ChatGPT, Claude and Google's AI a combined 2,961 times. It found a chance of less than 1 in 100 that ChatGPT or Google's AI would give the same list of brands in any two responses.
Perplexity wasn't in the study, so read it as a caution about measuring any of these engines, not as a Perplexity figure. A single run tells you little, however good the dashboard looks. A sound count reads "named in 12 of 40 answers", with the base beside it, and a gain of a few points on a small base is noise until repeat runs say otherwise.
Claims about llms.txt and domain rating aren't specific to Perplexity either, and Firemark's guide to getting cited by ChatGPT covers the evidence on both. A tip that comes with a precise percentage and no named study is a tip to skip.
How do you know a change worked?
Track the same buyer questions in Perplexity before and after a change, record named, cited and linked answers separately, and judge the result over repeat runs, not one.

A manual version takes four steps:
- Write 10 to 20 buyer questions and keep the wording fixed.
- Ask them in Perplexity and record, for each answer, whether your brand is named, whether your site is linked and which competitors appear.
- Make one change, such as opening the crawler or adding sourced statistics to a page.
- Ask the same questions again on later days, and compare counts over the same base.
Expect noise at every step. Answers can differ between two runs for reasons unrelated to your edit, so a jump after one change looks the same as a coincidence until the repeat runs agree. Record the date of each change, change one thing at a time, and leave a gap long enough for Perplexity's systems to reflect it, which Perplexity says can take up to 24 hours for its crawler settings.
Firemark sells a tracker that automates this, so apply the same scrutiny to it. Firemark tracks Perplexity on its Growth and Pro plans, and Lite and Starter don't include it. Each prompt goes to Perplexity's API with the sonar model and web search on, from a United States location in English, and is asked once per run. Because it reads the API, the Perplexity app can answer differently.
Firemark's method page states that it calls a movement only after three comparable runs and a change of 5 points or more. The tracker doesn't measure clicks or sales that follow an answer, so a change in your numbers shows what Perplexity says, not what readers do next.
Frequently asked questions
Does Perplexity use my content to train models?
Perplexity says it doesn't use either PerplexityBot or Perplexity-User to collect content for training foundation models. PerplexityBot exists to surface and link sites in search results, and Perplexity-User fetches a page when a user's question calls for it. Both statements come from Perplexity's own documentation, not from an outside audit.
Can I block Perplexity-User with robots.txt?
Perplexity says Perplexity-User generally ignores robots.txt rules because a user requested the fetch, so a robots.txt line isn't a dependable block. Perplexity publishes the fetcher's IP addresses, and any block would have to work at that level. PerplexityBot, the search crawler, is the one Perplexity asks you to allow in robots.txt.
Why is my page cited but my brand not named?
Perplexity can cite a page for one claim without naming the brand that published it, because retrieving a page, citing it and naming a brand are separate outcomes. Firemark counts named and linked answers separately for that reason, and a brand can be linked without being named. Read the answer text, not only the source list.
How often should you check Perplexity?
Check on a schedule you can compare, and judge movement over several runs rather than one. Firemark's method page calls a movement only after three comparable runs and a change of 5 points or more. Its guide to measuring AI visibility covers the setup.
Is Perplexity SEO the same as Google SEO?
Partly. In an Ahrefs study of short-tail terms published in September 2025, 65.07% of Perplexity's cited URLs were also in Google's top 10, so ordinary SEO may carry weight, and overlap was lower on long-tail prompts. Perplexity also builds each answer from retrieved pages and cites them by number, with no ranked list to hold, so the outcome to measure is different.
Measure Perplexity before you change anything
Perplexity has no rank to chase, so the first job is counting what it says about you. Firemark asks Perplexity your buyer questions and counts named and linked answers over a base, on its Growth and Pro plans. The Perplexity tracking page shows what's collected and what isn't. No step promises a mention, but a count tells you where you stand.





