Becoming discoverable in Perplexity: sources, crawlers and citations

Updated on July 29, 2026

Perplexity searches the web, compiles an answer, and displays numbered references to original sources. For companies, visibility therefore consists of several steps: the page must be accessible, relevant selected, correctly incorporated into the answer and, if necessary, visibly cited.

Channel Analysis | source optimization | periodic measurement

What is Perplexity?

Perplexity is an AI search engine that brings together web information into an instant answer with numbered citations and links to original sources.

The system can use different search modes and models. The concrete source selection and wording may therefore differ per question, mode, date, language and available web information. A single test does not provide a structural picture of your visibility.

What types of visibility should you distinguish?

Branding, source citation, and referral are different metrics and should not be reported as the same result.

How does citation work in Perplexity?

Perplexity uses web sources to support answers and displays numbered citations on web searches that allow users to open the original page.

A citation can support a specific claim, sub-topic or broader context. Therefore, do not assume that every single sentence is guaranteed to be linked one-to-one to one source. Always check which page is actually behind a reference and which part of the answer supports that source.

What crawler does Perplexity use for search results?

PerplexityBot is the official automatic crawler that can discover, index and link websites in Perplexity search results.

Perplexity notes that this crawler is not used to collect foundation model training content. To appear in Perplexity search results, you can allow PerplexityBot in robots.txt and prevent firewall or CDN rules from blocking the official crawler.

Which robots.txt line do you use for PerplexityBot?

Explicitly allow PerplexityBot when public website content may be available in Perplexity search results.

User-agent: PerplexityBot
Allow: /

Perplexity notes that changes to crawler settings may take up to approximately 24 hours for the systems to reflect them. A robots rule does not guarantee a crawl, inclusion, citation, or ranking.

What is Perplexity User?

Perplexity-User is a fetcher that can visit a page after a concrete user action in Perplexity.

It is not a regular automatic index crawler and is not used to collect content for foundation model training, according to Perplexity. Because the fetch was requested by a user, Perplexity-User typically ignores robots.txt. Anyone who wants to technically restrict access should therefore also review server, CDN and WAF rules.

What is the difference between PerplexityBot and Perplexity-User?

PerplexityBot automatically discovers content for search results, while Perplexity-User retrieves a page in response to a specific user query.

Should you allow Perplexity in your firewall?

Yes, when a Web Application Firewall or bot protection blocks legitimate requests and your public content is allowed to be accessible to Perplexity.

Perplexity recommends basing a WAF rule on the combination of the user agent and the officially published IP ranges. Simply relying on the name in the user agent is insufficient because that value can be impersonated.

• Use the current JSON endpoints from Perplexity as a source for the IP ranges.
• Link the correct IP list to PerplexityBot or Perplexity-User.
• Check if rate limits, bot management or challenge pages are blocking access.
• Give an allow rule sufficient priority compared to general blocks.
• Update IP sets periodically, as addresses can change.

A technical SEO analysis should also check rendering, canonicals, status codes, sitemaps and internal reachability.

How do you check if the crawler has access?

Check server, CDN and WAF logs and verify both the user agent and source IP.

1. Search logs for requests from PerplexityBot and Perplexity-User.
2. Compare the source IP to the correct official JSON endpoint.
3. Check status code, redirect chain and final canonical URL.
4. See if a challenge, login, consent layer or JavaScript dependency is blocking content.
5. Test key source pages using the same WAF and CDN configuration as production.
6. Recheck the logs after a robot or firewall change.

What helps a page to be useful as a resource?

A useful source page provides a specific answer, makes claims verifiable and clearly shows who is responsible for the content.

These are practical quality principles for strong web resources. Perplexity does not publish a complete list of website ranking factors that can guarantee inclusion or citation.

Does content need to be written specifically for Perplexity?

No. First and foremost, write a useful, public source for the real reader.

A channel-specific check is useful. Check whether PerplexityBot can reach the page, whether the main question is answered directly, whether the primary entity is clear and whether important claims can be verified. Avoid unnatural keyword variations or thin pages created only for one prompt.

How do you make passages quotable?

A quotable passage contains an independently understandable answer with sufficient context and a clear boundary to the claim.

• Start with the answer and then add explanations.
• Mention the subject or entity explicitly instead of just this or that.
• Use concrete concepts, periods and units.
• List conditions and exceptions in the same section.
• Link figures to measurement period, source and methodology.
• Use tables when comparing multiple options or features.
• Avoid marketing language that cannot be objectively verified.
• Don't put important information solely in images or widgets.

What role do entities and source architecture play?

A clear entity and source architecture makes visible which page contains the official information about an organization, person, service, location, product or case.

Use internal links to connect primary sources, supporting explanations, and evidence without any ambiguity. On Jakency link About Jakency, the profile of Jan Kenis and the projects and cases the same organizational and personal data with visible evidence.

Is structured data required for Perplexity?

No. Perplexity does not mention structured data in the official crawler and product documentation as an admission requirement or ranking factor for source selection.

Schema.org-markup can help to model the same Organization, Person, Service and other facts technically consistently. The visible HTML content remains leading. Don't put important claims exclusively in JSON-LD and don't make up Perplexity-specific schema types.

Do you need an llms.txt file?

No. Perplexity does not document llms.txt as a requirement for inclusion or citation of public websites.

A llms.txt file can be maintained as an experimental addition, but does not replace robots.txt, reachable HTML, clear navigation, canonicals, or strong source pages.

Do you have to constantly publish new content?

No. Current events are especially important when facts, prices, positions, people or rules actually change.

Update a page when the content requires it and make the measurement moment visible. Don't publish artificial changes just to show a new date. Perplexity emphasizes fresh web information as a product feature, but does not document a fixed freshness ranking factor for website owners.

How do you measure visibility in Perplexity?

Measure Perplexity separately with a fixed prompt set and record the brand mention, recommendation, citations, competitors, message and factual correctness per answer.

The full method is at Measure AI visibility. A GEO audit first records the initial situation. A GEO strategy then determines which demand clusters, channels, and source pages are prioritized.

Does a source citation prove that Perplexity generates traffic?

No. A source citation enables click-through, but traffic must be determined separately in analytics or server data.

• Track recognizable referral visits from Perplexity.
• Check landing page and corresponding answer theme.
• Measure engagement, follow-up and conversion beyond sessions.
• Please note that visits may be classified differently or incompletely.
• Report no-click brand mentions separately from referral traffic.

Do not make causal growth claims in analytics when only a citation or a few referrals are visible. Compare longer periods and record changes made in an action log.

What is the difference between ChatGPT Search and Google AI?

Perplexity, ChatGPT Search, and Google generative functions use web resources, but differ in crawler management, interface, resource representation, and measurability.

What work can Jakency carry out?

Jakency can analyze and improve technical access, source pages, entities and the measurement method around Perplexity.

View all GEO services or learn to perform the analysis internally via the GEO training.

What do you receive with a Perplexity analysis?

A Perplexity analysis provides a verifiable baseline measurement and a prioritized action plan.

What is not automatically included?

The analysis determines what needs to be improved, while the quotation states which implementation and external costs are included in the scope.

• Guarantee of inclusion, mention, recommendation or source citation.
• Complete rewrite and publication of all examined pages.
• Implementation of all technical, server and WAF adjustments.
• Continuous monitoring of all prompts, languages, markets and models.
• Modification of external websites or profiles without access or permission.
• Paid publications, digital PR, advertising or monitoring tools.
• Conversion, revenue or causal attribution without reliable data.

What does optimization for Perplexity cost?

The price depends on the number of prompts, pages, entities, markets, technical systems and desired measurement moments.

A detailed proposal follows after an initial conversation. A fixed package price without knowledge of the technical environment, question set and desired source pages would not be very reliable.

When is Perplexity optimization not the right first step?

Channel-specific optimization is less useful when the general information and SEO basis is not yet sufficiently reliable.

• Important pages are not public, crawlable or indexable.
• The organization has no clear services, entities or primary source pages.
• Claims lack authorship, evidence or current source information.
• There is no capacity to implement substantive or technical recommendations.
• The target group hardly uses Perplexity and other channels are demonstrably more important.
• The goal is only one positive citation without structural source quality.

What official Perplexity sources support this page?

Technical statements about crawlers, IP ranges, robots.txt and user-triggered fetches are based on current Perplexity documentation. The substantive optimization and measurement recommendations are practical advice from Jakency and are not presented as official or guaranteed Perplexity ranking factors.

• Perplexity Crawlers
• What is Perplexity?
• How does Perplexity work?
• Perplexity Search API