Perplexity SEO: how to become the answer, not just a result

Perplexity answers questions by reading a handful of pages and citing them inline. Being one of those citations is the whole game — there is no "position eleven." The good news: the selection process is more observable than classic Google ranking, because you can watch Perplexity's crawlers work.

Two visitors, two different jobs

Perplexity touches your site with two distinct agents, and conflating them leads to bad decisions:

• PerplexityBot builds the index — it decides whether you're eligible to be cited. Blocking it in robots.txt is de-listing yourself.
• Perplexity-User fires when a real person's question leads Perplexity to fetch your page live. Every one of these visits is a human reading you through an intermediary.

If you see the second in your logs, you're already being used as a source. The question is whether you're being cited — and clicked.

What actually gets cited

Working backwards from answers Perplexity gives in your niche, the cited pages share traits:

Direct answers near the top. Perplexity quotes passages, not pages. A page that opens with three paragraphs of throat-clearing before the answer loses to one that states the answer in the first hundred words and elaborates after.

One question per page. "Ultimate guides" that cover twelve subtopics dilute the match. Pages scoped to a single question — the kind people literally type — get pulled whole.

Verifiable specifics. Numbers, names, dates, steps. Vague content is hard to quote; concrete content quotes itself. (This is also why comparison tables and honest pros/cons lists earn citations — they're extractable.)

Fresh dates. Perplexity visibly prefers recently updated sources for anything time-sensitive. A true dateModified and content that matches it beat evergreen pretense.

Plain, crawlable HTML. Neither Perplexity agent executes your JavaScript app shell. If the answer only exists after hydration, it doesn't exist. (Same story we covered for Gemini — assistant fetchers read raw HTML.)

The measurement half nobody does

Most "Perplexity SEO" advice stops at content tips. The differentiating step is closing the loop:

1. Confirm you're being crawled. PerplexityBot in your traffic, verified against Perplexity's published IP ranges — not just claimed in a UA string.
2. Watch which pages Perplexity-User fetches. Those are the pages real users' questions route to. They're your citation surface — polish them first.
3. Track referral clicks from perplexity.ai. Most analytics buries these in Direct; they're recoverable. Rising fetches with flat referrals means you're being read but not cited compellingly — usually a first-paragraph problem.

That loop — crawled → fetched → cited → clicked — turns "Perplexity SEO" from folklore into a funnel you can actually optimize.

What not to do

Don't block PerplexityBot out of scraping anxiety while leaving Google full access — you're giving up the growing channel and keeping the crowded one. And don't stuff pages with "per Perplexity" phrasing or FAQ-schema spam; the selection favors pages that answer humans, and every observable signal says shortcuts age badly.

The uncomfortable summary: Perplexity rewards exactly the content quality Google always claimed to reward, but with a shorter feedback loop and fewer winners per query. Answer one question excellently per page, keep it in crawlable HTML, and measure whether the machine is actually reading you.