What Perplexity actually is
Perplexity is an AI-powered answer engine. The user asks a question; the system runs a web search, reads a small set of pages, and produces a synthesised answer with inline citations and a visible source list. The product is conversational like ChatGPT but oriented around explicit citations.
Two things matter for SEO. First, the citations are visible by default. Users see the source list every time. That makes Perplexity citations both more measurable (you can confirm whether you are being cited) and more clickable (the citation links are right there). Second, the user intent is unusually research-focused. People who reach Perplexity are usually verifying something, comparing sources, or digging into a topic. Click-through behaviour is correspondingly high-intent.
From an SEO perspective, Perplexity is the third-biggest AI search surface for most Australian businesses, behind AI Overviews and ChatGPT Search. Volume is smaller than either, but the citation visibility makes it the best surface for tracking and learning what kind of pages get picked up. The lessons from Perplexity citation tracking generally transfer to the other AI clients.
Why citation-first design matters
Perplexity's citation-first design is the single most useful feature for an SEO trying to learn what works. Two practical reasons.
- Easy verification. You can confirm whether your site is being cited for any query by running the query through Perplexity and inspecting the source list. The Sources panel shows every cited URL. No guessing, no inference.
- Citation-set learning. The full citation set is visible, not just the top one or two. That makes it easy to see which competitors are being cited alongside you, which authority sites are anchoring the set, and which patterns the system favours.
The practical implication. If you have time to track only one AI client manually each month, Perplexity is the highest-information one to track. The patterns you see there transfer broadly to ChatGPT Search and AI Overviews, both of which are harder to inspect at the same depth.
Perplexity's crawlers and robots.txt
Perplexity operates two distinct user agents.
PerplexityBot
The general indexing crawler. Fetches pages to build Perplexity's internal index. Allow or disallow via User-agent: PerplexityBot in robots.txt.
Perplexity-User
The live-search fetch agent. Used when a user's question requires fetching a current page. Allow or disallow via User-agent: Perplexity-User.
The recommended default for almost all marketing-focused Australian businesses: allow both. The content is already public and indexable by Google; allowing Perplexity to read it costs nothing and provides the visibility benefit of being available for citation. Disallowing PerplexityBot is reasonable for paid content sites where AI summarisation directly cannibalises the product, but disallowing Perplexity-User removes the site from Perplexity answers entirely, which is rarely the right move.
For the wider context on bot management, see the robots.txt chapter and the log file analysis chapter, which covers how to confirm in your access logs that the AI crawlers are actually visiting.
Signals we observe driving citations
Six patterns from monitoring Perplexity citation behaviour across client sites through 2025 and 2026. Observation, not algorithm.
Signal 1: General search visibility
The strongest predictor. Pages cited by Perplexity tend to rank well on major search engines (Google and Bing) for the underlying query. Sites with no organic visibility almost never appear in Perplexity citations. The work that wins Google rankings is also the work that gets you into Perplexity's candidate set.
Signal 2: Authority source clustering
Perplexity citations cluster heavily on a small number of authority sources per topic. Wikipedia, major publishers, government sites, established industry sites. Below that, the long tail is thin. Getting into the citation set requires either being the authority site or being the source whose content most directly answers the specific query.
Signal 3: Citable claims with named sources
Perplexity's synthesis step is visibly more comfortable lifting passages that contain named source attributions ("according to X", "the ABS reports", "a 2025 study by Y"). Pages with uncited assertions get summarised but rarely cited. The pattern matters most on factual claims; opinion content is less affected.
Signal 4: Named expert authors with Person schema
The E-E-A-T signal applies here as it does for the other AI clients. Articles with named credentialled authors and valid Person schema get cited more than anonymous content. The effect is strongest on YMYL topics. See E-E-A-T explained.
Signal 5: Direct-answer formatting
Pages that state the answer to the underlying question clearly, near the top, in a single declarative passage get extracted cleanly. Pages that bury the answer under preamble do not. The pattern is identical across all AI clients we have monitored.
Signal 6: External brand mentions
Sites with established external brand mentions (cited in industry publications, mentioned in major news, referenced in Wikipedia) tend to appear in Perplexity citation sets more often than equivalent sites without that external signal layer. The brand mentions chapter covers this work, and the digital PR chapter covers how to build the layer deliberately.
Practical optimisation moves
Five concrete steps for a site that wants to improve Perplexity citation pickup. As with ChatGPT Search, none of these are new SEO ideas; they are existing best practices applied more rigorously.
- Allow PerplexityBot and Perplexity-User in robots.txt. Confirm via manual robots.txt review. Re-check after any site migration.
- Win the underlying organic ranking first. Citation candidates are drawn from pages that rank. Pages on page two or lower are rarely candidates. The work to win the ranking is the existing SEO retainer.
- Tighten the entity layer.
Organizationschema withsameAs,Personschema for authors, transparent About page. Same move that wins everywhere else. - Add citable source attributions to factual claims. Where the page makes a factual assertion, attribute it to a named source. The synthesis step prefers lift-able citations.
- Track Perplexity citations monthly. Pick ten priority queries. Run them through Perplexity once a month. Record cited sources. The visible citation list makes Perplexity the easiest AI client to track manually.
Common mistakes
- Using Perplexity's visible citation list as the cheapest AI-search learning tool you have.
- Allowing both Perplexity user agents in robots.txt unless you have a specific opt-out reason.
- Adding named source attributions to factual claims.
- Investing in the brand mentions and digital PR layer that builds external authority.
- Treating Perplexity tracking as a monthly habit, not a one-off project.
- Blocking Perplexity crawlers reflexively. Most sites should allow them.
- Chasing Perplexity citations on pages that do not rank well anywhere else.
- Treating Perplexity as a separate optimisation programme from the other AI clients.
- Skipping the entity layer because "AI works it out". It does not.
- Trusting vendor claims of "guaranteed Perplexity citations". The algorithm is not public.
Perth and WA context
Two patterns from running Perplexity tracking across Perth and WA clients in 2026.
B2B and professional services see meaningful Perplexity exposure. Perth accounting firms, law firms, consulting practices and B2B services businesses get asked the kind of research-style questions Perplexity is built for. The work to win citations is the same E-E-A-T and entity-identity work that helps everywhere else. See legal SEO and the digital PR chapter for the authority layer that anchors Perplexity citation sets.
Local trade businesses see almost no Perplexity exposure. Perplexity is rarely used for "plumber near me" style queries; users go to Google Maps or the local pack. The work that matters for these businesses is the Local SEO pillar and the local citation layer, not Perplexity optimisation. See Local SEO Perth and trades SEO.
For the wider context, the AI Search pillar covers the full visibility stack, and the tracking AI referrals chapter covers how to measure the modest but high-intent traffic that Perplexity actually sends. For an entry-level diagnostic of the entity and authority signals that drive Perplexity pickup, the free SEO audit includes the relevant checks, and the full website audit service goes deeper.