9 July 2026

There is a question UK business owners have started asking that would have seemed absurd two years ago: “Why does Perplexity recommend my competitor when I search for our services — and what do I have to do to appear instead?” It is a legitimate and increasingly urgent question. Perplexity AI processed 780 million monthly queries in 2025, growing at a rate that has made it the fastest-adopted AI search product in the history of consumer technology. Unlike Google AI Overviews — which appear only for specific query types within Google’s own interface — Perplexity is where a growing segment of UK professionals, researchers, and high-intent buyers go first when they want a researched, synthesised answer rather than a list of links to wade through. The users asking Perplexity questions about which UK accounting software to use, which London SEO agency has genuine expertise in their sector, or which B2B SaaS platform integrates best with Salesforce are not casual browsers. They are people with specific needs, real budgets, and a demonstrated preference for trusted recommendations over sponsored results. Being cited by Perplexity in response to those queries is not a vanity metric. It is commercial visibility at the exact moment of highest purchase intent. This guide explains precisely how Perplexity decides which sources to cite, why the UK market represents a specific opportunity, and what UK businesses need to do — technically, editorially, and strategically — to become the sources Perplexity returns to consistently. How Perplexity Actually Works and Why It Is Different From Google Understanding Perplexity’s citation mechanics requires understanding that it is a fundamentally different product from Google, operating on a different retrieval and ranking model. Google indexes the web and ranks pages against queries using a complex algorithm built around backlinks, content relevance, technical signals, and hundreds of other factors. It serves links. The user reads the links and decides where to go. Perplexity is a retrieval-augmented generation (RAG) system. When a user submits a query, Perplexity runs a live web search — using its own crawler, Perplexitybot, alongside partnerships with search index providers — retrieves a set of candidate pages, passes those pages to a large language model (currently a combination of proprietary and third-party models including Claude, GPT-4 class models, and Mistral), and generates a synthesised answer from the retrieved content. It then cites the specific sources it drew from — typically two to five inline citations — directly within the answer. The critical distinction from Google is this: Perplexity does not rank pages and serve URLs. It reads pages and decides which ones contain the most credible, specific, extractable information for the query at hand. A page that ranks #1 on Google for a given query will not necessarily be cited by Perplexity — and a page that ranks #8 on Google might be cited repeatedly by Perplexity because its content structure, specificity, and credibility signals align more precisely with what Perplexity’s retrieval and generation system rewards. This means the optimisation signals for Perplexity citation are meaningfully different from traditional Google SEO — though they overlap significantly with the content quality principles covered in our AI Overviews, structured data, and prompt engineering posts in this series. Why the UK Market is Underserved in Perplexity’s Citation Pool There is a structural opportunity in the UK market that most UK businesses are not yet aware of, and most UK agencies have not yet acted on. Perplexity’s citation pool — the set of pages it draws from when generating answers — is dominated by US-origin content. This is partly a function of the web’s overall English-language content distribution (US publishers have historically produced the majority of English-language professional content), partly a function of Perplexity’s early user base being heavily US-weighted, and partly a function of the query types that drove Perplexity’s initial adoption being technology, finance, and startup topics where US content has deep authoritative coverage. For UK-specific queries — “best accountants for UK limited companies,” “GDPR compliance for UK SaaS businesses,” “how to register as self-employed UK,” “UK employment law redundancy process” — Perplexity frequently cites US-published content that addresses tangentially related but non-UK-specific topics, or cites UK government pages (.gov.uk) exclusively without citing any UK business or agency content. This is the gap. A UK business that produces genuinely authoritative, UK-specific content on topics where Perplexity currently has to reach for second-best sources is in a strong position to become one of Perplexity’s preferred UK citations — particularly if the content is structured for extraction, attributed to verified expert entities, and published on a domain with credible technical signals. The window to establish these citation positions is meaningfully wider in the UK market than it is in the US, where competition for AI citation in most professional service categories is already established. The Six Signals Perplexity Uses to Select Citation Sources Perplexity’s source selection is not fully transparent — its specific weighting of retrieval signals is proprietary. But systematic analysis of Perplexity answers across hundreds of UK business queries reveals consistent patterns in which sources are cited and why. Signal 1: Crawlability and technical accessibility. Perplexitybot must be able to crawl and index your pages. Unlike Google, which has accumulated a vast index over decades, Perplexitybot’s crawl coverage is more selective. Pages blocked by robots.txt, protected behind login walls, or returning slow server responses are effectively invisible to Perplexity’s retrieval system. The first step in any Perplexity optimisation programme is confirming that your site is crawlable by Perplexitybot — check your server logs for PerplexityBot user agent requests. If Perplexitybot is not in your logs, your content is not in Perplexity’s retrieval pool, regardless of how good it is. Signal 2: Semantic match to conversational query intent. Perplexity users submit conversational, often multi-part questions — not keyword fragments. “What are the main differences between sole trader and limited company for a UK freelancer in the tech industry?” is a typical Perplexity query. It is not “sole trader vs limited company UK.” Content that is