Nadia Mohamed
AI-SEO & GEO
How to Get Cited by Perplexity, and How to Check Whether It Already Cites You
To get cited by Perplexity, a page has to be reachable by PerplexityBot, match one of the searches Perplexity runs for the question, and be the kind of page Perplexity already prefers for that question type. For “best X” questions that usually means a ranked list; for “how do I” questions it means a page that answers the exact question in its opening lines.
This guide covers the crawler setup Perplexity documents, what Perplexity actually cited for seven prompts this site tracks in Ahrefs Brand Radar (two snapshots, a month apart), how product citations and product cards work, and how to check whether Perplexity cites you today.
Key takeaways
- Perplexity documents two agents:
PerplexityBot, which it says surfaces and links sites in its search results, andPerplexity-User, which fetches pages on a user’s behalf and generally ignores robots.txt. Allow the first, and whitelist both in your WAF if you use one. - On four “best X” prompts we track, 37 of the 40 pages Perplexity cited in September 2026 were ranked lists. For comparison questions, being listed on the pages Perplexity cites matters more than your own page.
- Citations churn: across seven tracked prompts, only 38 of the 70 URLs cited in August were cited again for the same prompt in September.
- A branded prompt can be lost to a namesake. Our “who is” prompt cited two of this site’s pages in August and none in September, when every citation went to a different person with the same name.
- Product citations have a separate route: Perplexity’s product cards draw on its index, Shopify data, a free merchant programme and PayPal’s catalogue sync, so product data matters as much as copy.
How Perplexity chooses what to cite
Perplexity was built to cite: it searches the web in real time and attaches clickable sources to its answers, so a citation is visible to the reader and can be tracked. It also runs its own index rather than borrowing Google’s or Bing’s. Ahrefs’ study of AI search overlap notes that Perplexity “has its own search index, based on its crawler”, PerplexityBot.
That index still behaves a lot like a search engine. Ahrefs found in August 2025 that around 12% of URLs cited by ChatGPT, Gemini and Copilot also appear in Google’s top 10, with Perplexity highest at 28.6% and ChatGPT’s in-text citations at 8.0%. The prompts were long-tail only, so head-term overlap is likely to be higher, and the study measured standalone assistants, not AI Overviews: in the same dataset, 76% of AI Overview citations came from top-10 pages.
Ahrefs’ explanation for the gap is query fan-out. Assistants turn one prompt into several searches and merge the results, so a page that ranks moderately for several variants can beat a page that ranks first for one. Ranking well helps with Perplexity, but for the searches it actually runs, which may not be the words the user typed.
The practical reading: treat Perplexity as a search engine with its own crawler, its own index and its own preferences for page types, and optimise for all three.
Step 1: make sure Perplexity can reach your pages
Perplexity’s crawler documentation lists two user agents, and they behave differently:
| User agent | What Perplexity says it does | Respects robots.txt? |
|---|---|---|
PerplexityBot | Surfaces and links websites in Perplexity’s search results. Perplexity says it is not used to crawl content for AI foundation models. | Yes |
Perplexity-User | Visits a page when a user asks a question, to help answer it and link to the page. | ”Generally ignores robots.txt rules”, because a user requested the fetch |
Perplexity recommends allowing PerplexityBot in robots.txt and permitting its published IP ranges. The documentation also says changes to robots.txt can take up to 24 hours to take effect. The minimal rule is:
User-agent: PerplexityBot
Allow: /
User-agent: Perplexity-User
Allow: /Robots.txt is only half of it. Firewalls and CDN bot rules block crawlers that robots.txt allows, and nothing in robots.txt will tell you. Perplexity’s documentation gives WAF recipes for Cloudflare and AWS: a custom rule that matches the user agent and the source IP against its published lists (perplexitybot.json and perplexity-user.json), with the action set to Allow and a higher priority than your blocking rules. Matching on both is the important part, because anyone can fake a user-agent string.
To confirm the rule works, check your access logs for PerplexityBot requests that come from the published ranges. The method, including IP verification, is in the AI crawler log analysis guide. If you are editing robots.txt for other AI crawlers at the same time, the robots.txt guide for developers covers the traps.
Two technical checks sit alongside crawler access:
- Server-rendered main content. If your headings and body text only appear after client-side JavaScript runs, check what a non-rendering fetch of the page actually receives. The JavaScript SEO guide shows how to test it.
- One canonical URL per page. Redirect duplicates and keep canonical tags consistent, so every citation goes to the URL you want. Our tracking below turned up a cited URL that now redirects.
What Perplexity cited for seven prompts we track
This site tracks seven custom prompts in Ahrefs Brand Radar, with Perplexity refreshed weekly and the market set to Germany. Two are branded (“Who is Nadia Mohamed and what does she do?” and a question about her approach to technical SEO and GEO). Five are unbranded: two about finding technical SEO consultants, one about a GEO and AEO expert for a SaaS launch in Europe, one about tools for tracking AI visibility, and one about how to tell whether ChatGPT, Perplexity, Gemini and Claude cite a brand.
We compared the answers dated 21 August 2026 with those dated 21 to 23 September 2026. That is 14 answers and 145 cited URLs. It is a small sample, and one set of prompts in one market, so read it as a worked example of what to look for, not a benchmark.
| Prompt type | Citations in Sept | Cited again from Aug | This site’s pages cited (Aug / Sept) |
|---|---|---|---|
| Branded: who is | 15 | 4 of 10 | 2 / 0 |
| Branded: approach | 10 | 5 of 10 | 5 / 3 |
| Unbranded: independent SEO consultants | 10 | 5 of 10 | 0 / 0 |
| Unbranded: B2B SaaS SEO consultants | 10 | 5 of 10 | 0 / 0 |
| Unbranded: AI visibility tools | 10 | 9 of 10 | 0 / 0 |
| Unbranded: GEO and AEO expert | 10 | 6 of 10 | 0 / 0 |
| Unbranded: is my brand cited | 10 | 4 of 10 | 0 / 0 |
| Total | 75 | 38 of 70 | 7 / 3 |
Five findings came out of it.
1. “Best X” questions are answered from ranked lists
On the four prompts that ask for the best consultants, agencies or tools, 37 of the 40 citations in September went to ranked-list pages: titles such as “Top 10”, “15 Best” or “The 18 Best … of 2026”. The three exceptions were a how-to article, one agency’s service page and a Reddit thread.
Only one of the 40 was a provider’s own service page. For a comparison query, the realistic route to a Perplexity citation is to be named on the lists Perplexity already cites, or to publish a genuinely useful list on your own domain. The tools prompt shows the second route working: LLM Pulse, whose own brand mention checker page was cited on another tracked prompt, had its “best AI visibility tools” roundup cited in both snapshots.
2. Citations churn month to month
Only 38 of the 70 URLs cited in August were cited again for the same prompt a month later. The AI visibility tools prompt was stable (9 of 10 kept), while the “is my brand cited” prompt kept only 4 of 10. A single check tells you very little. AI citation decay covers why citations erode and how to recover them.
3. A branded prompt can be taken by a namesake
In August, “Who is Nadia Mohamed” cited this site’s about page and a legacy author URL. In September, all 15 citations were about a different Nadia Mohamed, a US mayor, drawn from Wikipedia and news coverage. Perplexity searched for “Nadia Mohamed” and “Nadia Mohamed biography” and the more heavily covered entity won.
If your brand or name is shared, the bare name is not a query you own. Put the disambiguating context (role, company, topic) in page titles, in Organization schema and Person schema sameAs links, and in the way other sites describe you. The author entity guide covers the signals.
4. Branded questions cite hub pages, not articles
When the prompt named this site’s owner and topic, Perplexity cited the homepage, the insights index and a category page in September, plus two LinkedIn URLs. In August it cited the homepage, the insights index, the about page, the technical SEO consultant page and a legacy author URL. No individual article was cited in either month. For entity questions, the pages that describe who you are and what you do carry the citation, so keep them specific and up to date.
5. Perplexity cited a URL that now redirects
The August answers cited /author/nadia-mohamed/, an address from the site’s old WordPress build that now returns a 301 to /about/. The redirect still did its job, because the reader lands on the right page. But it shows Perplexity’s index can hold old URLs well after a move. After a migration, keep redirects in place and check which URLs Perplexity is actually citing.
Step 2: build pages that match the searches Perplexity runs
For six of the seven prompts, Brand Radar recorded Perplexity running a single search that was essentially the prompt itself. The branded “who is” prompt was the exception, rewritten into two shorter searches. So the question as a user phrases it is a good proxy for what you need to rank for.
That leads to three content rules:
- Match the page type to the question type. For “best”, “top” and “alternatives” prompts, the cited pages were lists. For “how do I” prompts, they were step-by-step guides. Writing a how-to for a comparison query is the most common mismatch.
- Answer the question in the opening lines. Make the first one or two sentences a complete answer that still makes sense if it is lifted out on its own.
- Keep the page current and dated. Many of the cited list pages carried the current year in their title. Update the substance, not just the date, and show a visible
updateddate that matches your Article schema’sdateModified.
Evidence helps a page stand out from the ten others Perplexity could cite. Link every statistic to its primary source, and publish first-party data where you have it. This page’s Brand Radar table is an example: nobody else can publish it, so it is the part most worth citing.
For a quick structural check of a draft, the AEO Article Analyzer scores an article against answer-first criteria and returns specific fixes.
How to increase product citations in Perplexity
Product questions reach your brand in two ways: as a cited source in the text answer, and as a product card.
Product cards
When Perplexity launched its shopping experience, its announcement said shopping questions return product cards alongside the answer, that the cards “aren’t sponsored”, and that discovery is powered partly by a Shopify integration covering Shopify-powered businesses that sell and ship to the US.
The same post introduced the Perplexity Merchant Program, free for merchants and aimed at large retailers who share product specs. Perplexity says members get an increased chance of being a “recommended product”, because the products are in its index with fuller details. Other benefits it lists are inclusion in its one-click checkout, free API access and a dashboard of search and shopping trends. Sign-up is through a form linked from the post.
In November 2025, PayPal announced that PayPal merchants could become discoverable inside Perplexity. Its “store sync” service makes product catalogues discoverable in Perplexity’s results, and US users can check out in the chat with the retailer as merchant of record. Perplexity’s own shopping post describes the experience as free to all US users and says it uses a shopper’s past searches to personalise results.
What that means in practice:
- Get the catalogue into Perplexity’s data sources. Depending on your stack, that means the Shopify integration, PayPal store sync or the Merchant Program. All three are US-first according to the announcements above.
- Make product pages complete and crawlable. Put specs, price, availability and variants in server-rendered HTML that PerplexityBot can read. Perplexity says fuller details help it judge whether a product is relevant.
- Expect personalisation. Perplexity says results use each shopper’s history, so two people asking the same question can see different cards. Track trends across many prompts rather than a single screenshot.
Cited sources on product questions
The text part of a product answer follows the same pattern as our “best X” data: Perplexity cites roundups, reviews and comparison pages. To be cited on those questions, get the product onto the review and list pages that already rank for your category, and publish comparison content that is honest enough to be cited. A product page that only describes itself is rarely what a “which should I buy” answer links to.
How to check whether Perplexity cites you
Perplexity’s crawler documentation covers access, not reporting, so you have to measure citations yourself. Use three methods together.
1. A fixed prompt set, checked on a schedule
Write 10 to 20 prompts that match how your buyers ask: branded, category (“best X for Y”) and problem (“how do I …”). Run the same set on a fixed schedule and record, for each answer, whether your domain is cited, which URL, and which domains are cited instead. The manual prompt matrix sets this up step by step.
Given the churn in our data, repeat each prompt more than once before drawing conclusions. The AI share of voice method explains how many responses you need before a change means anything.
2. Referral traffic in GA4
Perplexity clicks arrive as referrals from perplexity.ai. A custom channel group for AI sources separates them from other referrals, and the landing-page report shows which pages Perplexity sends people to. The setup is in how to track AI referral traffic in GA4.
Referral data shows clicks, not citations, and conversion claims for this traffic rest on thin evidence. In Seer Interactive’s case study of a single client between October 2024 and April 2025, ChatGPT referral traffic converted at 15.9% against 1.76% for Google Organic on the same site; Perplexity 10.5%, Claude 5%, Gemini 3%. That is one client, with just under 11,000 AI sessions against almost 14 million from Google, and Seer’s own title is “6 Learnings, 1 site”. It is a case study, not a benchmark, and it is now more than a year old.
3. A monitoring tool, once manual checks stop scaling
Tools run your prompt set on a schedule and store the cited URLs. The Brand Radar data in this article came from one of them. Two reference points:
- Otterly.ai. Lite is $29 a month for 15 prompts, Standard is $189 for 100 and Premium is $489 for 400, each tracking ChatGPT, Google AI Overviews, Perplexity and Microsoft Copilot (read from Otterly’s pricing page on 24 September 2026).
- Peec AI. The Peec AI review covers what it tracks, what it costs and who it suits.
For a wider comparison, see the best AI visibility tools. Whichever tool you choose, check that it stores the full list of cited URLs for each answer and lets you set the country, because both shaped the findings above.
How Perplexity differs from ChatGPT
Perplexity aims to support nearly every statement with a source, while ChatGPT decides query by query whether to search at all, according to Ahrefs’ overlap study. A page that ranks well for the searches Perplexity runs therefore has a more predictable path to a Perplexity citation than to a ChatGPT one. Perplexity vs ChatGPT search compares the two engines in detail, and how to get cited by ChatGPT covers the other side.
If you want the crawler access, entity signals and page structure above audited and implemented across your priority pages, that is the work of a GEO and technical SEO engagement.
Frequently asked questions
How do I know if PerplexityBot can crawl my site?
Check robots.txt for a User-agent: PerplexityBot group that allows your content, then look in your server logs for PerplexityBot requests from the IP ranges Perplexity publishes. If robots.txt allows it and there are no verified requests, a WAF or CDN bot rule is the likely block. Perplexity’s documentation gives allow-rule recipes for Cloudflare and AWS WAF that match both the user agent and the IP range.
Does Perplexity only cite pages that rank in Google?
No, but it overlaps with Google more than other assistants do. In Ahrefs’ August 2025 long-tail study, 28.6% of Perplexity’s citations ranked in Google’s top 10 for the same prompt, the highest of the assistants measured, against around 12% on average. That study used long-tail prompts only and measured standalone assistants, not AI Overviews. Perplexity runs its own index and may search for variants of the prompt, so pages that rank for related searches can be cited too.
Why does Perplexity cite a competitor’s list instead of my product page?
For “best X” questions, Perplexity mostly cites ranked lists: 37 of 40 citations on the four such prompts this site tracks. Your product page answers “what is X”, not “which X should I choose”. Getting named on the lists Perplexity already cites, or publishing an honest comparison of your own, fits that pattern better than optimising the product page for the query.
How can I get my products into Perplexity’s product cards?
Perplexity says its shopping results draw on its own index, a Shopify integration and its free Merchant Program for large retailers. Since November 2025, PayPal merchants can also sync their catalogues so products are discoverable in Perplexity. Whichever route you use, keep specs, price and availability complete and crawlable, because Perplexity says fuller product details help it judge relevance.
How often should I check Perplexity citations?
Monthly is a sensible minimum for a fixed prompt set, with each prompt run more than once. In our tracking, only 38 of the 70 URLs cited in August were cited again for the same prompt a month later, so a single run can easily mislead. Weekly tracking through a tool makes it easier to separate a trend from noise.
Can I pay Perplexity to be cited?
There is no paid route into the organic citations. Perplexity says its shopping product cards are not sponsored and that its Merchant Program is free and separate from its advertising products. Citations come from being crawlable, relevant to the searches Perplexity runs, and the page type it prefers for the question.