How to get cited in Perplexity: the UK research guide

Perplexity is the AI engine UK marketers most underestimate. It has a fraction of ChatGPT's user base, but the users it does have are disproportionately high-intent (researchers, analysts, B2B buyers), and the citation model is the most transparent of any major AI engine. If you want to learn how AI search ranking works, optimising for Perplexity is the fastest classroom.

This is what 800+ tracked queries across UK client engagements have taught us.

How Perplexity actually retrieves sources

Perplexity performs a live web search for every query (no training-data fallback), retrieves 5-15 candidate sources, and synthesises an answer that explicitly cites a subset of them. The citations are visible and clickable. This makes Perplexity uniquely measurable: you can directly see whether you were cited, and if so, in which position.

The retrieval substrate is a hybrid: Perplexity's own crawl, plus the Bing index for breadth. The selection model is more recency-weighted and more authority-weighted than ChatGPT Search.

The signals that move Perplexity citations

In order of observed impact in our test set:

1. **Topical authority at the domain level.** Perplexity favours domains that have published consistently on a topic over time. A single great article on a generalist site loses to a moderate article on a specialist site. This is the most asymmetric signal in AI search and the one most UK brands ignore. 2. **Recency.** Articles updated within 60 days are heavily favoured on commercial and current-events queries. We have watched citations shift within 48 hours of a republish. 3. **Source diversity in your own corpus.** Perplexity de-duplicates by domain in many answer sets. Having multiple distinct pages targeting the query family (a pillar plus 4-6 supporting articles) increases the probability that at least one is cited. 4. **External corroboration.** Same authority signal as ChatGPT, weighted slightly higher in Perplexity. Brands cited in trade publications and industry directories appear in Perplexity citations significantly more often. 5. **Schema and structured data.** Less critical for Perplexity than for ChatGPT Search, but FAQPage and Article schema still help. 6. **Backlinks.** The only AI engine where backlinks show a measurable correlation with citation rate. Not the dominant signal, but present.

The 4-week Perplexity playbook

**Week 1: Probe.** Run your 30 priority queries through Perplexity. Document citation presence, position and competitor citations. The free AVO Visibility Index covers Perplexity in the default probe set.

**Week 2: Publish or refresh.** For each query where you're absent, identify the strongest existing article on your site and update it with a 60-day-fresh date, a tighter summary paragraph, a comparison table where applicable, and 2-3 new sourced statistics. Where no article exists, publish one.

**Week 3: Corroborate.** Pitch one guest article or podcast appearance per priority query family on a Perplexity-indexed publication. Trade press in your category is more valuable than mass-market press.

**Week 4: Re-probe and iterate.** Re-run the queries. In our test set, 35-50% of refreshed articles gained citation within four weeks. The remainder typically need an authority intervention rather than a content intervention.

Why Perplexity matters strategically

Three reasons. First, the measurement loop is fast enough to learn from. Second, Perplexity's user base over-indexes on buyers who write the briefs that hire you. Third, the signals that win Perplexity citations generalise. Brands that perform well in Perplexity tend to perform well in ChatGPT Search and Google AI Overviews within 2-3 months, because the underlying signals (topical depth, recency, corroboration) are shared.

If you want one AI engine to optimise for first, this is the one.