Does Perplexity pay publishers?

Perplexity has run a publisher programme with revenue sharing for selected partners. It is invitation-based and does not change how ordinary sites are chosen as sources.

Does Perplexity use Google or Bing?

It operates its own index via PerplexityBot and has not published a full list of any other providers it uses. Treat its own crawler access as the part you control.

Why do Reddit threads outrank us?

Forum threads carry first-hand answers in plain language, which match conversational questions closely. Answering the same question plainly on your own page, and in the thread, covers both routes.

Engine guide: Perplexity

How to get cited by Perplexity

Perplexity is the one assistant that searches every time and puts its sources at the top of the answer. That makes it the easiest to test and the least forgiving of a slow or blocked page.

The short answer

Perplexity retrieves live for nearly every question and cites numbered sources above and inside the answer. It runs its own crawler, PerplexityBot, for its index, and a separate Perplexity-User agent for fetches a person triggers. Allow PerplexityBot, keep pages fast and server-rendered, and state facts with dates, because it favours recent, specific passages.

What ranked on 26 September 2026: The ranking pages are mostly GEO tool blogs and a handful of case studies. Perplexity's own crawler documentation rarely appears on page one, although it is the only primary source for which user agents to allow.

What makes Perplexity different

Every answer is grounded, so the question is never whether it will search, only which pages it will pick. It also shows more sources per answer than most assistants, which lowers the bar to appear at all.

PerplexityBot builds Perplexity's index and is documented as respecting robots.txt; it is not used to train foundation models
Perplexity-User fetches pages a person asks about in a thread, and Perplexity says it generally ignores robots.txt because a human requested the page
Sources display as numbered cards before the answer text, so a title and a clean domain affect whether anyone clicks
It leans on recent material for news, product and pricing questions, which rewards visible dates and changelogs
Focus modes such as academic or social narrow the source pool, so the same brand can appear in one mode and vanish in another

Things that sound right and are not

ClaimBlocking PerplexityBot keeps our content out.
It keeps you out of the index and out of most citations. User-triggered fetches can still read a page someone pastes in. Decide what you want before you block, because you cannot block your way into being cited.
ClaimLong guides get cited more.
Perplexity quotes passages, not pages. A long guide with the answer in paragraph fourteen loses to a short page that answers in its opening lines. Length helps only when each section stands on its own.
ClaimOne test run tells us where we stand.
Source selection shifts between runs, modes and models chosen in the settings. Repeat each prompt several times and across modes before drawing a conclusion.

Where to start

01Allow PerplexityBot in robots.txt and in any bot-management rules at the CDN.
02Check that key pages return full HTML quickly without client-side rendering.
03Add a visible last-updated date and a short summary block to pages that answer time-sensitive questions.
04Run your core prompts in default and relevant focus modes, and note which source domains recur.
05Get your details correct on the recurring third-party domains, since those are the pages Perplexity already trusts for your topic.

Questions

Does Perplexity pay publishers?: Perplexity has run a publisher programme with revenue sharing for selected partners. It is invitation-based and does not change how ordinary sites are chosen as sources.
Does Perplexity use Google or Bing?: It operates its own index via PerplexityBot and has not published a full list of any other providers it uses. Treat its own crawler access as the part you control.
Why do Reddit threads outrank us?: Forum threads carry first-hand answers in plain language, which match conversational questions closely. Answering the same question plainly on your own page, and in the thread, covers both routes.

Doing this yourself and want a second opinion on what the assistants say about you? Tell me what you are seeing, or read the AEO Playbook for the whole framework.