How to Get Cited by Perplexity (Measured Playbook)

Last verified: 2026-08-11. This playbook comes from running a fixed query set against Perplexity every week and recording which sources it cites — including a month where our own project had perfect on-page metadata and zero mentions.

The finding that changes everything

Perplexity does not crawl GitHub and sort by stars. In our measured category (open-source GEO tools), projects with 17–37 stars appeared in answers while a project with 147 stars did not — for over a month. The difference was not product quality or README optimization. It was presence in the sources Perplexity synthesizes from:

• Aggregators (LibHunt was the top-cited source in our category).
• Curated lists (awesome-* repositories with their own authority).
• Niche blogs and comparison articles with dates and concrete criteria.

If those sources don’t mention you, Perplexity doesn’t either.

Step 1 — Map what Perplexity cites for your queries

1. Write down 8–10 fixed queries your buyers ask (category, comparison, how-to).
2. Run them weekly. Record the cited domains, not just whether you appear.
3. The cited domains are your distribution target list. This is observable, not speculative — one query run shows you exactly which curators matter in your niche.

Step 2 — Get into those sources

• Submit to the aggregators and directories that appear in answers.
• Contribute honest entries to curated lists (read each list’s inclusion criteria first; a silently-closed PR is a signal to open an issue asking for criteria, not to spam).
• Publish the kind of third-party-style content yourself: dated comparisons that credit competitors. Pages with a visible “Last verified” date and a stated methodology are easier for an engine to trust and cite.

Step 3 — Keep PerplexityBot unblocked

User-agent: PerplexityBot
Allow: /

Check your CDN/WAF separately — robots.txt permission means nothing if a bot-fighting rule blocks the fetch upstream.

Step 4 — Publish citable pages on your own domain

Perplexity cites pages, not repositories. Each target query deserves a page with:

• a direct answer in the first 50–170 words;
• a stated methodology (“how we built this comparison”);
• a visible verification date;
• Article or FAQPage JSON-LD;
• links to primary sources.

Step 5 — Entity consistency

Use one identical description of your product across your site, GitHub, package registries, and every listing. Answer engines build entity confidence from repetition across independent surfaces; drift between descriptions reads as uncertainty.

What our measurements look like

We track 10 fixed queries (2 branded, 8 generic) weekly against Perplexity and snapshot the results. The metric that matters is % of the fixed set mentioning you, plus the cited-domain list. One run is a sample; the trend over weeks is the signal.

---

In short: Perplexity visibility is earned off your own site first — get into the curator sources it cites, keep PerplexityBot unblocked, publish dated citable pages, and measure weekly with a fixed query set. E-GEO implements this loop as open-source tooling (fixed query sets, snapshot diffs, content optimization): github.com/mverab/eGEOagents.

Related: How to rank in ChatGPT search — the same measurement loop against OAI-SearchBot and Bing’s index.