Products SEO & GEO Agent Paid Ads Agent Website Agent Conversion Agent
How it works Pricing FAQ Blog Book a call

Perplexity citations: what it cites and what makes a page citable

Updated · 7 min read

Short answer Perplexity cites pages from its own web index, scored section by section. Across all topics, Reddit, YouTube and Wikipedia get the most citations. For local questions, business websites and directories such as Yelp, MapQuest and trade directories lead. Your page has a chance when one section answers one question with specific facts and Perplexity's crawler can read it.

Want help applying this to your business? Talk it through on a free 30-minute call.

Book a call

What sources does Perplexity cite most?

Forums, video and reference sites. Ahrefs tracked every site Perplexity cited across more than 3.1 million US questions on all topics and ranked them by mention share: each site's slice of the citations that the 50 most-cited sites earn together. November 2026 snapshot:

Type of sourceMost-cited sites (mention share)
ForumsReddit 24.6%
VideoYouTube 14.2%
EncyclopediaWikipedia 7.2%
StoresAmazon 4.3%, Alibaba 1.7%, Target 1.3%
Social networksFacebook 3.1%, LinkedIn 1.1%, Instagram 1.0%
Health publishersHealthline 2.9%, WebMD 1.4%
Reviews and directoriesTripadvisor 1.6%, Yelp 1.1%, Trustpilot 1.0%

That top 50 is a thin slice of the whole. Surfer counted about 10.37 million Perplexity citations from questions its users tracked in September 2026. The 50 most-cited domains got 9.0% of them; the other 91.0% went to sites outside that list. Surfer's users lean toward marketing topics, so its top 50 differs from Ahrefs', with Reddit first in both. For a question about your trade in your town, most cited pages are ones no national ranking shows.

What does Perplexity cite for local business questions?

Business websites for fact questions, directories for "best" questions. Yext, which sells listing management, studied 6.8 million citations from ChatGPT, Gemini and Perplexity in July and August 2025. It asked through each model's API, at the locations of its clients and prospects in finance, food, healthcare and retail:

Kind of questionWhat got cited
A fact, no business named ("[service] near me")Business websites: over 40% of citations, in all three models
An opinion, no business named ("best [service] in [city]")Industry-specific directories: 24% of Perplexity's citations, the highest of the three
Listings and reviews, all questionsPerplexity favored MapQuest (over 364,000 citations) and Tripadvisor (over 239,000)

A smaller BrightLocal test in July 2025 (20 searches in 10 industries, on four AI tools) agrees. Perplexity used Yelp in every industry tested, cited MapQuest often, and leaned on Super Lawyers and FindLaw for legal questions. Tripadvisor's reviews also reach Perplexity through a partnership the two announced on January 9, 2025.

So there are two ways in. Your own page answers the fact questions. The directories of your trade answer the "best" ones.

How does Perplexity choose which page to cite?

It searches its own index and scores pieces of pages, not only whole pages. Perplexity's research team described the system in September 2025; the same search stack runs the Perplexity app:

  • Its own index of over 200 billion URLs, filled by its crawler, PerplexityBot, and by third-party crawlers it partners with (Help Center: How does Perplexity follow robots.txt?).
  • Words and meaning. It looks up pages both by the words they use and by what they mean, then merges the two lists.
  • Sections, not just pages. It cuts each page into sections that stand on their own and scores each one:
"Having well-structured, comprehensive parsing allows us to decompose documents into self-contained spans, each of which can be individually retrieved and ranked at query time." (Perplexity Research: Architecting and evaluating an AI-first Search API)
  • Filters first. Early stages drop "clearly non-responsive or stale content".
  • Authority and layout. Pages from "authoritative domains" get priority in fast storage, and sites heavy on data "in list or table form" may get more set reading rules.

Perplexity doesn't publish what makes a domain authoritative. It does tag some cited sites as Government, Academic or Trusted. Its review rates whole sites with questions such as whether a site corrects its mistakes and says who wrote each piece. A missing tag means little:

"Most domains on the web do not have a label. No label does not mean a site is low quality, and it is not a negative judgment." (Perplexity Help Center: Understanding source labels)

It doesn't say whether tags change which page gets cited.

What makes a page citable?

A section that answers one question by itself, with facts that can be checked. Each step, and the reason behind it:

Do thisWhy
Open each section with the answer, naming the service, the place and the numberPerplexity scores sections one by one
Head each section with the customer's question, in their words ("How much does tile roof repair cost in Phoenix?")It matches by words as well as meaning
Put prices, service areas, hours and licenses in page text, lists or tablesIts reader may use set rules for list and table data
Name the author and correct errors openlyIts source review asks about both
Update facts when they change, and the date with themIt drops stale content. Old pages still get cited: Ahrefs (July 2025) found Perplexity's cited pages averaged 1,166 days since publication, against 1,416 for Google's results
Let PerplexityBot inPerplexity recommends it "to ensure your site appears in search results"; a blocked page may keep only "the domain, headline, and a brief factual summary" in its index
Complete the directories that show up for your questionsFor "best" questions, trade directories got 24% of Perplexity's citations (Yext)

Crawler setup: how to rank in Perplexity, Gemini and Claude. Directories: local citations and directories.

What doesn't Perplexity say it uses?

Some common advice has no Perplexity source behind it:

  • Schema markup (code that labels facts for machines). Perplexity's research post describes reading the page's content and layout, and doesn't mention schema. Google says its own AI features need none: local business schema markup.
  • Your bounce rate or time on page. Perplexity says its ranking learns from its own users' requests and human feedback on its answers. It doesn't mention your site's analytics.
  • Backlinks (links from other sites). It names "authoritative domains" but not how it judges them.

How do you see which sources Perplexity cites for you?

Ask it what your customers ask, and read the sources:

  1. Write 10 questions people ask before they hire you: 5 facts ("who does same-day furnace repair in Denver?") and 5 opinions ("best furnace repair in Denver").
  2. Ask each in Perplexity, with your city in the question.
  3. Open the sources under each answer. Note every domain: your site, competitors, directories, Reddit threads, videos.
  4. Circle the domains that repeat. Fix your listing on each one, or write the page that answers the fact question better than what's cited.
  5. Repeat monthly; answers change. The method matches the 20-question ChatGPT test.

Each cited source also has a "Rate this source" option for feedback on that domain (Help Center: Understanding source labels).

Example

Illustrative cases. Numbers are invented for the example.

A Phoenix roofer (US$)

"Best tile roof repair in Phoenix" cites Yelp, BBB, a Reddit thread and two competitors' service pages. His tile page is a photo gallery with one line of text. He adds a section that answers the cost question ($500 to $1,500 for most repairs), lists the suburbs he serves and shows his license number. Then he completes his Yelp and BBB profiles.

A Calgary moving company (C$)

"How much does a two-bedroom move cost in Calgary?" cites a Reddit thread and a competitor's price table. The owner publishes a dated table of crew size, typical hours and price by home size (two-bedroom: C$900 to C$1,400), plus what raises the price. Each row answers a question by itself.

An Austin estate planning lawyer

"Best estate planning lawyer in Austin" cites Super Lawyers, FindLaw and Avvo; her site never appears. She completes those profiles within her bar's advertising rules. For fact questions such as "do I need a will if I have a trust?", she writes one short page per question.

FAQ

Can I pay Perplexity to cite my business?

Perplexity describes no such program. Its paid deals are with publishers, such as Gannett (July 2025) and Comet Plus (August 2025).

Should I post on Reddit to get cited?

Reddit tops both the Ahrefs and Surfer rankings. Posting there means following each community's rules on self-promotion: Reddit for a local business.

Is there a Perplexity business listing I can claim?

Perplexity's docs describe none. In the studies above, its local citations came from business websites, MapQuest, Yelp, Tripadvisor and trade directories. Keep those current.

Is it different in Canada?

Perplexity's help pages describe none. The Ahrefs figures are US questions, so run the 10-question check with your Canadian city and fix the directories that appear.

Different for you?

Keep learning

NextPerplexity traffic: who it sends and how to measure itHow much traffic Perplexity sends a small business, who those visitors are, where Google Analytics 4 files them, and how to count what it misses.

Do it yourself

Does ChatGPT recommend your business? A 20-question testA free test any owner can run to see whether ChatGPT names their business, which competitors it names instead, and why.

Ask us one question

One line is enough. We reply by email.

Sources

  1. Perplexity Research: Architecting and evaluating an AI-first Search API (September 25, 2025)
  2. Perplexity Help Center: Understanding source labels
  3. Perplexity Help Center: How does Perplexity follow robots.txt?
  4. Perplexity Docs: Perplexity crawlers
  5. Perplexity: Welcoming Gannett to the Perplexity Publisher Program (July 30, 2025)
  6. Perplexity: Introducing Comet Plus (August 25, 2025)
  7. Tripadvisor Group: Tripadvisor Group and Perplexity partner (January 9, 2025)
  8. Ahrefs: The 50 Most-Cited Websites in Perplexity (updated November 3, 2026)
  9. Surfer: The 50 Most-Cited Domains in Perplexity (September 2026)
  10. Yext Research: AI Citations, User Locations, and Query Context (October 2025)
  11. BrightLocal: AI Search Makes Local Listings More Important Than Ever (July 22, 2025)
  12. Ahrefs: New Study: AI Assistants Prefer to Cite Fresher Content (July 28, 2025)