---
title: "How to Get Cited by Perplexity: What the Evidence Actually Shows"
description: "Perplexity cites sources that directly answer the query, are reachable by its crawler, and carry checkable data — but independent research shows domain authority and backlinks barely predict who gets picked, and there's no guaranteed formula."
answer_summary: "Perplexity cites sources that directly answer the query, are reachable by its crawler, and carry checkable data — but independent research shows domain authority and backlinks barely predict who gets picked, and there's no guaranteed formula."
canonical: "https://nqz.ai/blog/geo-how-to-get-cited-by-perplexity"
published_at: "2026-08-11T05:18:34.957Z"
updated_at: "2026-08-21T07:37:43.000Z"
author: "Ada O'Brien"
category: "GEO"
tags: ["geo","aeo","ai-search","perplexity","ai-citations"]
image: "https://images.unsplash.com/photo-1633356122544-f134324a6cee?w=1200&h=630&fit=crop"
---

# How to Get Cited by Perplexity: What the Evidence Actually Shows

Perplexity cites pages that directly answer the query it's synthesizing, aren't blocked from its crawler, and contain checkable facts it can attribute — but which specific page wins that slot is decided by a retrieval-and-ranking system Perplexity hasn't fully disclosed, and the best independent data available shows several popular "optimization" tactics have little to no measurable effect.

This piece separates what Perplexity has said officially, what independent researchers have measured, and what's still marketing speculation dressed up as strategy.

## What "getting cited by Perplexity" actually means


**Direct answer:** Perplexity is a retrieval-augmented generation (RAG) product: it turns a query into several search sub-queries, pulls candidate pages, and has a language model write an answer that attributes specific claims to specific URLs. A "citation" is a numbered footnote or inline link pointing at one of those source pages. Unlike classic SEO, there's no fixed ranking position to chase — a page either gets pulled into the answer and cited, or it doesn't, and the criteria are query-specific rather than site-wide.


Two things make this harder to game than traditional search:

- **There's no submission channel.** Perplexity doesn't offer a webmaster portal for pitching content, unlike Google Search Console. Discovery is entirely through its own crawler and third-party indexes.
- **Citation is per-answer, not per-site.** A domain can be cited constantly for one topic and never for an adjacent one. Aggregate "domain authority" is a weak proxy for whether any given page gets picked.

## How Perplexity actually decides what to cite


**Direct answer:** **Crawlability is a hard prerequisite.** Perplexity operates two separate crawlers with different rules, according to its own developer documentation: `PerplexityBot` indexes pages for search results and respects `robots.txt`, while `Perplexity-User` fetches pages on behalf of a live user typing a question and generally ignores `robots.txt`, the same way a human clicking a link would ([Perplexity Crawlers docs](https://docs.perplexity.ai/docs/resources/perplexity-crawlers)). Perplexity states neither crawler is used to train its underlying models. If `PerplexityBot` is blocked, your indexed content can't be cited from search — that part is unambiguous and within your control.


Whether it's actually followed reliably is disputed. In August 2025, Cloudflare published data showing what it called "stealth, undeclared crawlers" tied to Perplexity — traffic that rotated user agents (including one impersonating Chrome on macOS) and shifted across IP ranges after being blocked, generating what Cloudflare estimated at 3–6 million daily requests against sites that had explicitly disallowed crawling. Cloudflare de-listed Perplexity as a verified bot as a result. Perplexity disputed the attribution, saying the traffic came from a third-party browser-automation service (Browserbase) responding to individual user requests, not its own crawler ([Cloudflare, "Perplexity is using stealth, undeclared crawlers..."](https://blog.cloudflare.com/perplexity-is-using-stealth-undeclared-crawlers-to-evade-website-no-crawl-directives/)). The honest takeaway: robots.txt compliance is Perplexity's stated policy for its declared crawler, but the boundary between "crawler" and "user-triggered fetch" has been a genuine, unresolved point of conflict — not settled fact.

**What actually correlates with getting picked.** The most rigorous independent test of this comes from Ahrefs, which compared Ahrefs Rank (a backlink-based authority score) against how often the top 50 most-cited domains in Ahrefs' Brand Radar were mentioned by Perplexity across roughly 953,500 real prompts in June 2025. The result: a Spearman correlation of -0.34 (p = 0.095) — a weak relationship, and in the wrong direction to support "more backlinks = more citations" ([Ahrefs, "Domain-Level Link Metrics May Not Be Good Predictors of AI Search Mentions"](https://ahrefs.com/blog/domain-link-metrics-not-good-mention-predictors/)). A separate Ahrefs analysis of 15,000 long-tail prompts found only 28.6% of Perplexity's citations pointed to pages that also ranked in Google's top 10 for the same query — the highest overlap of any AI assistant tested, but still meaning more than 7 in 10 of Perplexity's citations go to pages that aren't top-10 Google results at all ([Ahrefs, "Only 12% of AI Cited URLs Rank in Google's Top 10"](https://ahrefs.com/blog/ai-search-overlap/)). Translation: ranking well on Google helps somewhat but is far from sufficient, and classic link-building is a weak lever specifically for Perplexity.

**What Perplexity actually cites, by the numbers.** In that same June 2025 Ahrefs dataset, Perplexity's most-cited domains were YouTube (16.1% of citations) and Wikipedia (12.5%), a pattern distinct from both ChatGPT and Google AI Overviews:

| Platform | Sample size (Jun 2025) | Top domain | 2nd | 3rd |
|---|---|---|---|---|
| Perplexity | ~953,500 prompts | YouTube — 16.1% | Wikipedia — 12.5% | Apple News/AP |
| ChatGPT | ~957,000 prompts | Wikipedia — 16.3% | Reuters — 4.0% | Apple News/AP — 2.6% |
| Google AI Overviews | ~76.7M searches | YouTube — 9.5% | Wikipedia — 8.4% | Reddit — 7.4% |

*Source: [Ahrefs, "The 10 Most Mentioned Domains for ChatGPT, Perplexity, and AI Overviews"](https://ahrefs.com/blog/top-10-most-cited-domains-ai-assistants/), June 2025.*

Separately, an academic study of news citations specifically — analyzing over 366,000 citations across roughly 65,000 responses from OpenAI, Perplexity, and Google systems on the AI Search Arena platform — found citation patterns are heavily concentrated among a small number of outlets in every system, though which outlets differ by provider ([Kai-Cheng Yang, "News Source Citing Patterns in AI Search Systems," arXiv:2507.05301, July 2025](https://arxiv.org/abs/2507.05301)). If you're not already one of the handful of dominant sources in your topic's news category, don't expect frequent citation on breaking-news-style queries; niche, non-news informational content operates under different, less concentrated dynamics.

**What independently-tested content changes actually move the needle.** The most direct evidence on content-level tactics comes from the original "Generative Engine Optimization" (GEO) paper — a peer-reviewed study from Princeton, Georgia Tech, the Allen Institute for AI, and IIT Delhi, presented at KDD 2024. Testing nine content interventions across a diverse query set, the researchers found that adding statistics, direct quotations, and citations to sources produced the largest, most consistent visibility gains (up to roughly 40% in combination), while keyword stuffing — the classic SEO tactic — actually reduced visibility by about 10% ([Aggarwal et al., "GEO: Generative Engine Optimization," arXiv:2311.09735, KDD 2024](https://arxiv.org/abs/2311.09735)). This is one of the only controlled, reproducible tests in this space — most other "ranking factor" claims you'll read are pattern-matching from observational data, not causal tests.

## What's genuinely unproven

**Direct answer:** Be skeptical of anything stated as fact in this list — none of it has independent, controlled evidence behind it as of this writing:

- **"Content updated in the last 30 days gets a citation boost."** Repeated across SEO blogs with no cited methodology or sample size. Plausible given RAG systems generally favor fresher results, but not independently measured for Perplexity specifically.
- **"Perplexity pulls 5–10 pages and cites 3–4 per query."** A commonly repeated figure with no traceable primary source. Treat as folklore, not fact.
- **Schema markup directly causing citation.** Structured data helps machines parse your page, but no public study ties FAQPage/HowTo schema to a measurable Perplexity citation lift.
- **Submitting your site to Perplexity for consideration.** There is no such mechanism. Discovery is entirely crawler- and index-driven.

## A practical checklist

1. **Confirm you're not blocking `PerplexityBot`.** Check your `robots.txt` and any WAF rules against the documented user agent string. This is the only item on this list Perplexity has stated outright as a prerequisite.
2. **Answer the specific question in the first sentence or two of the relevant section.** RAG systems extract and attribute short, self-contained claims more reliably than they extract meaning buried in long paragraphs.
3. **Back claims with numbers, direct quotes, and named sources — and cite them yourself.** This is the one tactic with controlled-study evidence behind it (GEO paper, above).
4. **Don't rely on backlink building as your primary lever for this specific platform.** The Ahrefs correlation data says domain authority is a weak predictor for Perplexity relative to other signals.
5. **Don't assume ranking #1 on Google guarantees a Perplexity citation, or that failing to rank there rules you out.** Under 30% overlap either way.
6. **Check manually.** There's no analytics dashboard for "how often Perplexity cited me." Run your target queries yourself, periodically, or use a monitoring tool built for this (more below).

## The honest limits

There is no guaranteed way to get cited by Perplexity, and anyone selling one is overstating what's known. Perplexity doesn't publish its ranking weights, the algorithm changes without notice, and the company is in active legal disputes with publishers — Dow Jones and the New York Post sued Perplexity in October 2024, alleging it reproduces copyrighted news content in answers, sometimes without adequate attribution ([CNBC, "Murdoch's Dow Jones, New York Post sue Perplexity AI," Oct. 21, 2024](https://www.cnbc.com/2024/10/21/murdoch-firms-dow-jones-and-new-york-post-sue-perplexity-ai.html)) — which means how Perplexity handles publisher content is itself still being contested in court, not fixed policy. In August 2025, Perplexity responded to publisher pressure by launching Comet Plus, a $5/month subscription tier that pools 80% of its revenue for participating publishers based on how often their content is surfaced or visited by Perplexity's agents ([Engadget, "Perplexity has cooked up a new way to pay publishers for their content," Aug. 2025](https://www.engadget.com/ai/perplexity-has-cooked-up-a-new-way-to-pay-publishers-for-their-content-204255019.html)) — a sign the underlying citation and compensation mechanics are still being actively redesigned, not a stable target to optimize against long-term.

## Where nqzai fits

nqzai doesn't have a tool that gets you cited by Perplexity specifically — no vendor legitimately does, for the reasons above. What it does have: [AI search visibility tracking](https://nqz.ai/ai-search-optimization) that runs your target prompts against ChatGPT, Claude, Gemini, and Perplexity plus Google AI Overviews, and reports which specific URL or domain got cited as the source — so instead of guessing whether you're showing up, you can see it directly. Paired with that, the [content quality and E-E-A-T checks](https://nqz.ai/content-optimization-ai-tools) score a page against a documented rubric and flag the structural gaps (missing direct answers, no checkable data, weak sourcing) most correlated with the tactics that actually have evidence behind them above. It's a monitoring and diagnostic layer, not a promise.

## FAQ

**Does Perplexity crawl and index the whole web like Google?**
No. Perplexity uses its own indexing plus outside search APIs, and it's explicit that neither of its crawlers — `PerplexityBot` or `Perplexity-User` — is used to pre-train its underlying AI models. Its index is smaller and more selective than Google's general web index.

**Can I submit my site directly to Perplexity to get considered for citations?**
No. There's no publisher submission portal. Getting crawled and potentially cited depends entirely on `PerplexityBot` being able to access your pages and the retrieval system judging them relevant to a given query.

**Does blocking PerplexityBot in robots.txt hurt my citation chances?**
Yes, directly. Perplexity's own documentation states `PerplexityBot` respects `robots.txt`, so disallowing it removes you from consideration for that crawler's index. Separately, `Perplexity-User` (live, user-triggered fetches) generally doesn't check `robots.txt`, and Cloudflare has published evidence disputing how consistently Perplexity's declared crawler behaves under blocking — so treat "blocked = fully invisible" as directionally true but not airtight.

**Do backlinks help you get cited by Perplexity?**
Weakly, at best, according to the only controlled correlation study available: Ahrefs found a -0.34 Spearman correlation between domain-level link authority and Perplexity mention share across a 50-domain sample — meaning backlinks are a much weaker predictor here than they are for classic Google rankings.

**How is Perplexity different from ChatGPT or Google AI Overviews in what it cites?**
Perplexity's top-cited domains skew toward YouTube and Wikipedia; ChatGPT leans more heavily on Wikipedia and wire services like Reuters; Google AI Overviews spreads more broadly and includes Reddit in its top three. All three concentrate citations among a relatively small set of dominant domains, especially for news queries.

**Is there a guaranteed way to get cited by Perplexity?**
No. Perplexity doesn't publish its ranking criteria, they change without notice, and even the best-documented lever (adding statistics, quotes, and citations, per the peer-reviewed GEO study) produced gains in aggregate testing, not guaranteed results for any single page or query.