Pre-flight checks
- robots.txt allows PerplexityBot and Perplexity-User.
- Fetch test: key pages return full content to a plain HTTP fetch, fast, no JS dependency, no bot-protection false positives.
- Harvest baseline: run your 30-prompt set, record every cited URL.
Source priorities
- Owned pages — answer blocks, tables, honest dates. Perplexity lifts structure mechanically; the best-shaped chunk wins regardless of domain fame.
- Niche authority pages — the specialized guide out-cites the famous homepage here.
- Freshness cadence — evolving topics reward recent, genuinely-updated pages.
30-day plan
| Week | Moves | Checkpoint |
|---|---|---|
| 1 | Fetchability fixes; answer blocks on top-10 pages | Pages fetch clean; dates visible |
| 2 | Comparison tables and spec passages on money pages | First inline citations appear |
| 3 | Two corroborating third-party placements | Cited-URL inventory diversifies |
| 4 | Re-run prompt set; analyze which passages got lifted | Citation Rate movement documented |
Perplexity-specific notes
- Use it as your research instrument: the inline citations on your competitors' winning answers are a free, continuously-updated map of your category's source graph.
- Comet and API surfaces: citation increasingly happens inside the browser and embedded contexts — the same fetchability rules apply.
Frequently asked questions
Why is Perplexity where you show POC results first?
Because it has no memory layer and no index lag — retrieval is live and citation is transparent. A fix that ships Monday can be a citation by Friday. It's the cleanest before/after evidence in GEO.