Content Strategy

How to Optimise for ChatGPT and Perplexity

6 min read

Why these two platforms matter most right now

ChatGPT and Perplexity are currently the two highest-traffic AI search platforms that actively retrieve and cite web content. Google AI Overviews draws from the existing Google index, so optimising for it is largely an extension of existing SEO. ChatGPT and Perplexity use separate retrieval systems that require deliberate attention.

Perplexity is the more citation-visible of the two: it shows source links prominently in every answer and drives measurable referral traffic to cited pages. ChatGPT browsing mode cites sources less consistently, but its user base is large enough that appearing in its answers is commercially significant even without reliable traffic attribution.

How Perplexity retrieves and ranks content

Perplexity performs a live web search for most queries. It retrieves a set of candidate pages, reads their content, and synthesises an answer with inline citations. The pages it selects are those it judges most likely to contain a direct, accurate answer to the specific query.

  • It checks for llms.txt: pages on sites with a well-formed llms.txt are given additional context about the publisher, which improves how Perplexity represents the content.
  • It favours pages with clear, early answers: content where the first paragraph addresses the query directly is extracted more reliably than content that builds to an answer.
  • It uses your robots.txt: PerplexityBot must be allowed. If your robots.txt blocks it under a wildcard rule, Perplexity cannot index your content at all.
  • It weights freshness: recently published or recently updated pages are preferred for queries where the answer may change over time. Keep dateModified accurate in your Article schema.
  • It values concise, factual sentences: Perplexity often pulls a single sentence from a page to use as its citation. Sentences that state a fact clearly, with no hedging, are the ones that get pulled.

How ChatGPT retrieves and ranks content

ChatGPT uses web browsing in two modes: it searches proactively when a query is likely to need current information, and it searches on user request when a user asks it to look something up. In both cases it retrieves pages and incorporates their content into its response.

  • GPTBot and OAI-SearchBot must be allowed in your robots.txt: ChatGPT uses both for indexing and live retrieval. Blocking either reduces your presence in its answers.
  • It favours authoritative pages: ChatGPT weights publisher authority heavily. Organization schema with sameAs links to established profiles, and Article schema with named authors, both contribute to how it evaluates source quality.
  • It prefers structured content: pages with clear H2 headings, short paragraphs, and explicit answers are easier to incorporate into a coherent response. Long, essay-style content with no visual structure is harder to extract cleanly.
  • It uses OpenAI's search index: content that OAI-SearchBot has indexed is available for retrieval without a live fetch. Keeping your content crawlable and your sitemap current helps ensure your pages are in the index.

The changes with the most impact across both platforms

  • Allow the right crawlers: PerplexityBot, GPTBot, and OAI-SearchBot must all be explicitly allowed in your robots.txt. This is the prerequisite for everything else.
  • Add or fix Article schema: include author with sameAs, datePublished, dateModified, and publisher with sameAs. Both platforms use this to evaluate source authority.
  • Write a direct first paragraph: for every page you want cited, the first paragraph should answer the page's primary question. This is the single structural change with the most consistent effect across both platforms.
  • Add llms.txt: primarily for Perplexity, but worth doing for all AI platforms. It gives them context about your site as a whole rather than requiring them to infer it from individual pages.
  • Keep content fresh: update dateModified whenever you make substantive changes to a page. Both platforms use freshness signals for time-sensitive queries.
You do not need to write different content for each AI platform. The same changes that improve your Perplexity citability improve your ChatGPT citability. The foundation is identical: crawler access, schema markup, direct answers, and a verified publisher identity.

What to check this week

  • Navigate to yoursite.com/robots.txt and confirm PerplexityBot, GPTBot, and OAI-SearchBot are explicitly allowed.
  • Check your top 10 pages for Article schema with author and dateModified populated.
  • Pick your three most-trafficked pages and verify the first paragraph answers the page's primary question directly.
  • Navigate to yoursite.com/llms.txt. If you get a 404, create the file.

The SEOFliq AEO and GEO Suite extension scores your page against 8 AI platforms including ChatGPT and Perplexity readiness, audits your robots.txt for all major AI crawlers, and checks for llms.txt. It runs on any page in seconds with no account required.