Perplexity AI and ChatGPT Search: How to Get Your Site Referenced in 2026

Two years ago, Perplexity AI and ChatGPT Search were novelties with negligible traffic impact. In 2026, they are significant referral traffic sources for sites that have been selected as citation sources. Perplexity AI exceeded 100 million monthly searches in 2025. ChatGPT Search, integrated into ChatGPT Plus with its hundreds of millions of users, has introduced web search capabilities to the largest AI user base in the world. Both platforms cite their sources, which means getting cited sends actual, measurable traffic to your website. This guide covers exactly how each platform selects sources, what optimisations increase your citation probability, and how to track whether you are appearing in AI search results.

Perplexity AI is a dedicated AI search engine built around the concept of fully cited answers. Every response Perplexity generates includes numbered citations with visible URLs. The platform uses a combination of its own web crawler (PerplexityBot) and integration with Bing's index to find and retrieve source content at query time. This real-time retrieval approach means Perplexity can cite content published within hours or days — it is not limited to a periodically updated model training dataset.

Perplexity's source selection algorithm prioritises content that is: directly relevant to the specific query, easy to parse (the full content is accessible in the HTML response without JavaScript execution), from a domain with established subject-matter authority, and structured in a way that makes specific claims extractable. Perplexity tends to favour specialist sites over generalist sites for specific topics — a dedicated cybersecurity blog will consistently outperform a general technology news site for cybersecurity queries, even if the general site has higher overall domain authority.

ChatGPT Search, launched by OpenAI in late 2024, integrates live web search into ChatGPT's conversation interface. When a ChatGPT query requires current or real-time information, the system performs a web search and retrieves source content, which it incorporates into its response along with citations. OpenAI uses its own crawlers (GPTBot and OAI-SearchBot) for its web index, supplemented by Bing's index through a partnership that informed the initial launch.

ChatGPT Search's source selection is more strongly correlated with Bing's organic ranking signals than Perplexity's is, because of the Bing index partnership. Pages that rank well in Bing are more likely to appear as ChatGPT Search citations, particularly for commercial and informational queries. This gives Bing SEO — historically neglected — a concrete traffic multiplier: strong Bing rankings feed directly into ChatGPT Search citation visibility.

Both platforms use named crawlers to index content. If these crawlers are blocked in your robots.txt, your site cannot be cited by the respective platform. Check your robots.txt now and verify that the following user agents are not blocked:

• PerplexityBot — Perplexity AI's primary crawler. User agent string: 'PerplexityBot'
• GPTBot — OpenAI's web crawler for ChatGPT's training and search. User agent string: 'GPTBot'
• OAI-SearchBot — OpenAI's search-specific crawler, distinct from GPTBot. User agent string: 'OAI-SearchBot'
• Bingbot — Microsoft's crawler, which feeds ChatGPT Search through Bing integration. User agent string: 'Bingbot'

In 2024, many SEO guides recommended blocking GPTBot to protect content from AI training. This advice conflated training data crawls with search-time retrieval crawls. Blocking GPTBot prevents your content from being used in OpenAI's model training, but it also blocks real-time search retrieval for ChatGPT Search. The two use cases cannot currently be separated — blocking GPTBot blocks both. Sites that blocked GPTBot based on 2024 guidance and want ChatGPT Search visibility need to re-evaluate that decision.

Perplexity's retrieval system has several quirks that inform specific optimisation decisions beyond general AEO principles.

• Server-side rendering is strongly preferred. Perplexity's crawler has limited JavaScript execution capability. Pages that require JavaScript to render the main body content are frequently crawled as near-empty pages. If your site uses client-side rendering (React SPA, Vue SPA), ensure critical content is available in the initial HTML response through SSR or SSG.
• Clean HTML structure matters more than visual design. Perplexity parses semantic HTML — h1, h2, h3, p, ul, ol, strong, em — and uses this structure to identify the hierarchy of information. Complex CSS layouts, overuse of div nesting, and semantic HTML violations (e.g., using strong for decoration rather than emphasis) make parsing harder.
• Freshly published content is accessible quickly. Because Perplexity performs real-time crawls at query time, recently published content can be cited within days of publication if it is relevant to active query topics. For news and trending topics, publishing quickly and ensuring the page is crawlable immediately is a Perplexity-specific timing advantage.
• Heading structure is the primary navigation signal. Use the Heading Analyser at /tools/heading-structure-analyser to audit your page's H1–H3 hierarchy. Perplexity's system uses headings to identify content sections, and a clear, hierarchical heading structure directly improves the system's ability to extract and cite specific sections.
• Avoid interstitials and aggressive consent banners. Pages that show a full-screen consent banner before the content is accessible may be crawled as near-empty. If your consent management solution blocks page rendering until consent is granted, configure it to allow crawlers to access the page content without triggering the consent flow.

Unlike Google Search Console, neither Perplexity nor ChatGPT currently provides a publisher dashboard showing when and how often your content is cited. Tracking AI search citation requires indirect methods.

• GA4 referral traffic. In Google Analytics 4 or Ahrefs Web Analytics, filter your Acquisition > Traffic Acquisition report for 'perplexity.ai' as the referral source. Sessions from Perplexity citation clicks appear as referral traffic from this domain. ChatGPT Search clicks appear as referral traffic from 'chat.openai.com'. Track this weekly — a growing referral count from these sources is direct evidence of increasing AI citation visibility.
• Manual query testing. Search your target queries directly in Perplexity and ChatGPT. Note which sources are cited. If competitors are cited for your target queries and you are not, compare your content structure and crawlability to theirs. This is time-intensive but provides direct intelligence about citation patterns for your specific topics.
• Server access logs. Your web server logs show all crawler visits, including PerplexityBot and GPTBot. Reviewing access logs to see which pages these crawlers are visiting — and how frequently — gives an early signal of crawl coverage before citation data is available.
• Third-party AI visibility tracking tools. Services including Profound, Semrush, and Ahrefs have added or are adding AI search visibility tracking that monitors citation frequency across Perplexity and ChatGPT Search at scale. These tools are more reliable than manual spot-checking for tracking trends across large keyword sets.

In 2025, Perplexity launched 'Perplexity Pages' — a feature that allows users to create shareable, formatted summaries of Perplexity research results as public pages indexed on perplexity.ai. These pages cite their sources and are publicly accessible. For sites that become consistent citation sources for Perplexity Pages on popular topics, the compound effect is significant: each Perplexity Page that cites your content becomes an additional source of referral traffic and an additional backlink on a high-domain-authority platform.

Approach AI search citation the same way you approach building organic search authority — as a medium-term investment in content quality and topical depth, measured through consistent monitoring rather than immediate results. The sites that are gaining the most AI search traffic in 2026 started investing in high-quality, structured, citation-ready content in 2023 and 2024. Their investment is now generating compounding returns as more users search through AI platforms.

The specific actions that build AI search presence: ensuring all crawlers have access (robots.txt audit), using server-side rendering or SSG, optimising heading structure with the Heading Analyser at /tools/heading-structure-analyser, writing answer-first paragraphs, adding FAQ schema to informational pages, publishing into a focused topic area consistently, and building external links from authoritative sources within your topic. None of these actions are unique to AI search — they are all good content and SEO practices that happen to be precisely what AI search citation requires.