Perplexity Optimization Checklist

Perplexity is the most citation-transparent answer engine — every answer ships with numbered source links — which makes it both the easiest engine to optimize for and the easiest to measure. This checklist covers the four things that determine whether Perplexity cites you: crawler access, freshness, literal-query retrievability, and monitoring. Work through it once, then rerun the freshness and monitoring sections quarterly.

Crawler access: can Perplexity reach you at all?

• Allow PerplexityBot in robots.txt — this is the index-building crawler; blocking it removes you from Perplexity's corpus entirely.
• Allow Perplexity-User in robots.txt — this agent fetches pages live when a user's question triggers real-time retrieval, so it produces citations even for freshly published URLs.
• Test both user agents through your CDN with curl -A "PerplexityBot" -I https://yoursite.com — Cloudflare's "Block AI Scrapers" toggle and similar WAF rules override robots.txt silently.
• Confirm both bots appear in server logs within a week — absence after seven days means a block is still active somewhere in the stack.
• Serve full content in the initial HTML response (SSR or SSG) — PerplexityBot does not execute JavaScript, so client-rendered content is invisible to it.

Freshness: Perplexity's strongest ranking bias

• Display a visible "Last updated" date near the top of every money page — Perplexity weights recency more aggressively than any other major engine.
• Keep dateModified in your Article schema synchronized with the visible date — mismatches read as manipulation.
• Refresh statistics pages at least quarterly with genuinely new numbers — republishing unchanged content with a new date is detected and penalized.
• Ping IndexNow after every meaningful update — instant re-crawl signals beat waiting for a scheduled visit.
• Maintain a refresh queue ranked by citation value — pages already earning Perplexity citations deserve refresh priority because incumbency plus freshness compounds.

Retrievability: match the literal question

• Title pages with the question phrasing users actually type — Perplexity retrieval rewards close lexical matches to the query.
• Answer the title's question completely in the first 2-4 sentences — the opening paragraph is the passage most likely to be lifted into the answer.
• Use question-shaped H2s for every subtopic — each heading is an independent retrieval hook for adjacent queries.
• Keep passages self-contained at 40-80 words — Perplexity quotes chunks, not pages, and a chunk that references "as mentioned above" loses its meaning.
• Add at least three concrete facts per page (numbers, dates, named sources) — the GEO research by Aggarwal et al. (KDD 2024) measured a 30-40% generative visibility lift from adding statistics and citations, and Perplexity's answer style visibly favors quantified sources.
• Put comparisons in real GFM/HTML tables — Perplexity lifts tables into answers wholesale.

Monitoring: prove the checklist worked

• Build a tracked prompt set of 20-50 questions your buyers ask, phrased naturally — this is your baseline for AI share of voice.
• Record which domains occupy Perplexity's citation slots for each prompt weekly — citation tracking automates this sweep so you see wins and losses instead of anecdotes.
• Segment perplexity.ai referrer traffic in analytics — Perplexity sends real clickthroughs, unlike some engines, so referral volume is a direct outcome metric.
• Re-test your priority prompts within a week of any content refresh — freshness effects on Perplexity show up fast, which makes it the best engine for validating whether your GEO changes actually move citations.

The compounding loop matters more than any single item: access enables indexing, freshness wins the retrieval tiebreak, structure wins the citation slot, and monitoring tells you which pages to refresh next. Teams that run the freshness and monitoring sections on a fixed quarterly cadence consistently outperform teams that treat Perplexity as a one-time setup task.