Short answer
- Perplexity searches the web in real time and shows numbered sources. To be one of them,
PerplexityBotmust be allowed; Perplexity recommends it and says it is not used for training foundation models. - Perplexity cites community sources heavily: Reddit was its #1 cited domain in Ahrefs’ September 2026 data.
- Pages that answer one question clearly, with dates and specifics, are the easiest to cite.
How Perplexity chooses sources
Perplexity answers by searching the web at query time and attaching numbered source footnotes. It documents two agents (Perplexity crawler documentation):
| Agent | What it does | Control |
|---|---|---|
PerplexityBot | Surfaces and links websites in Perplexity search results; Perplexity states it is not used to crawl content for AI foundation models | robots.txt; Perplexity recommends allowing it. Changes can take up to 24 hours |
Perplexity-User | Fetches a page when a user’s question needs it; the page may be linked in the answer | Perplexity says it generally ignores robots.txt, because a user requested it |
Perplexity does not publish how its index ranks sources. Its access practices have also been disputed: in August 2025 Cloudflare alleged undeclared crawling, which Perplexity rejected. What is observable is the citation pattern: Perplexity shows many sources per answer and favours pages that state a specific answer near the top.
Reddit’s role in Perplexity answers
In Ahrefs Brand Radar data for September 2026, reddit.com was the #1 cited domain in Perplexity answers, at 21.6% of citations among the top-50 sources (Ahrefs). Shares move: Semrush tracked ChatGPT search citing Reddit in close to 60% of responses in early August 2025 and about 10% by mid-September 2025 (Semrush). Treat any single number as a snapshot. Partnerships give engines structured access to Reddit content: Google (Feb 2024) and OpenAI (May 2024).
What this means in practice: when buyers ask “best X” or “is X worth it”, the answer often quotes a real discussion. You cannot write that discussion yourself, but you can make sure your brand is represented accurately in it:
- Find the threads already cited for your prompts and read what people actually say about you and competitors.
- Answer questions where you have real expertise, from a clearly labelled brand or employee account.
- Follow each subreddit’s self-promotion rules; ask moderators before an AMA, a launch post or a giveaway.
- Correct factual errors politely with a source; never argue with reviewers.
- Never use vote manipulation, undisclosed accounts or paid posts disguised as users. They break Reddit’s rules and tend to be removed, which also removes the citation.
This is the work of our Reddit community marketing service: subreddit research, rules-first participation and moderator cooperation.
10-point diagnostic checklist
- robots.txt and CDN/WAF rules allow
PerplexityBot. Test it free with our AI crawler checker. - Each important question has one page that answers it in the first paragraph.
- Numbers, dates and named sources are in the text (Perplexity tends to quote specifics).
- Pages show a visible updated date, and old facts are corrected.
- You have comparison and “alternatives” pages for your category.
- You know which Reddit threads, forums and review pages Perplexity cites for your prompts.
- Your brand participates in those threads openly and within each community’s rules.
- Third-party listings (directories, review sites, marketplaces) carry the same facts as your site.
- Analytics segments referrals from
perplexity.ai. - You run the same prompts monthly and log which URLs are cited.
Want this done for you? Our GEO service runs the checklist, fixes your pages and works on the sources Perplexity cites, including rules-first Reddit community marketing.
How to measure Perplexity visibility
We found no official Perplexity analytics tool for publishers. Track referrals from perplexity.ai and keep a monthly prompt log with the cited URLs. If citations drop suddenly, check bot rules first; see our Perplexity visibility drop diagnostic.
How Perplexity compares with other AI engines
| Engine | Where answers come from (official) | Crawler / control to check | Reddit in top cited sources* | How to measure |
|---|---|---|---|---|
| ChatGPT search | Live web search via OpenAI’s search crawler plus partner content; OpenAI does not publicly name a single search provider | Allow OAI-SearchBot (search). GPTBot is training only | #1 · 16.8% | Referrals tagged utm_source=chatgpt.com; prompt tracking |
| Perplexity | Real-time web search with numbered source footnotes | Allow PerplexityBot (Perplexity recommends it) | #1 · 21.6% | perplexity.ai referrals; prompt tracking |
| Google AI Overviews | Google Search index and core ranking systems, with query fan-out | Googlebot, snippet controls, Search Console AI opt-out | #2 · 18.5% | Search Console Performance (Web) and generative AI report |
| Google AI Mode | Same index as Search; may use different models, so links differ from AI Overviews | Same as AI Overviews | #1 · 17.9% | Same as AI Overviews |
| Gemini | Some responses grounded on Google Search results | Google-Extended controls Gemini training and grounding use | #1 · 28.5% | Prompt tracking (no publisher report found) |
| Microsoft Copilot | Copilot sends a short query to Bing and composes from Bing results | Bing indexing; IndexNow for freshness | Not in top 50 | Bing Webmaster Tools AI Performance (citations) |
*Ahrefs Brand Radar “mention share”, US, September 2026: Reddit’s share of citations among each engine’s top-50 cited domains, not of all citations. Sources: ChatGPT, Perplexity, Google AI Overviews, Google AI Mode, Gemini, Microsoft Copilot.
Related guides
FAQ
How do I get my website cited by Perplexity?
Allow PerplexityBot in robots.txt and CDN rules, publish pages that answer a specific question with concrete facts near the top, keep them dated and current, and be represented accurately on the community and review pages Perplexity cites for your topic.
Does PerplexityBot use my content to train AI models?
Perplexity states that PerplexityBot is used to surface and link websites in search results and is not used to crawl content for AI foundation models.
Can I block Perplexity from fetching my pages?
You can disallow PerplexityBot in robots.txt, which removes you from its search index. Perplexity says Perplexity-User, which fetches pages a user asks for, generally ignores robots.txt.
Why does Perplexity cite Reddit so much?
Perplexity favours sources with first-hand detail, and Reddit threads often contain it. In Ahrefs Brand Radar data for September 2026, Reddit was the most cited domain in Perplexity answers among the top-50 sources.
See how Perplexity describes your brand today, and which sources it cites instead of you.
Get a free GeoTrack diagnosis