TL;DR: ChatGPT and Perplexity retrieve sources differently, cite from different pools, and share only 11% of domains in common (Averi, 680M citation study, March 2026). Getting cited by both requires understanding what each system selects and the structural signals that cut across both. Almost no neutral guide on this exists. This is one.
Why does "ranking in ChatGPT" mean something different than ranking in Google?
Answer capsule: "Ranking in ChatGPT" is not ranking. Google returns a list of URLs. ChatGPT and Perplexity retrieve content, select passages, synthesize an answer, and credit a small number of sources. The question is not whether your page ranks. It is whether your content gets retrieved, selected, and cited. These are three separate filters, each with distinct signals.
Most businesses searching for how to rank in ChatGPT or Perplexity are asking the wrong question. Google ranks pages. ChatGPT and Perplexity cite sources. The distinction changes everything about what you optimize.
In Google, the goal is position. A page that ranks #1 gets roughly 30% of clicks. A page that ranks #10 gets roughly 1%. The ranking system is the product.
In AI search, position doesn't exist. ChatGPT retrieves 16+ URLs per query and cites roughly half of what it retrieves from search index sources (Ahrefs, 1.4M prompt study, February 2025). Perplexity evaluates 60+ candidate sources per query and assembles an answer with citations embedded before generation (ZipTie.dev, 2025–2026). Your page either gets selected and credited, or it doesn't appear at all.
Three separate filters decide the outcome: retrieval (does the system find your page?), selection (does it pass quality and relevance gates?), citation (does it appear in the final answer?). Failing any one of them means your page contributed to the answer without being credited, which is common and invisible without proper tracking.
Understanding what generative engine optimization actually means starts here.
How does ChatGPT decide what to cite?
Answer capsule: ChatGPT retrieves sources through Bing's index and real-time crawl. When a search-enabled query triggers retrieval, sources from the search index get cited at 88.46%, the highest of any retrieval channel (Ahrefs, 1.4M prompt study, February 2025). Content structure matters more than domain authority at the selection stage. Natural language URL slugs get cited at 89.78% vs. 81.11% for opaque URLs in the same study.
ChatGPT operates in two distinct modes, and the distinction matters.
In standard chat without web search enabled, responses draw on training data. For most business queries, this means the system answers from its pre-training corpus, which has a knowledge cutoff and underrepresents newer, smaller domains by design. Being cited in this mode requires being in the training data in a recognizable, authoritative form.
In search-enabled mode (ChatGPT Search, mobile, and most current desktop queries), the system queries Bing's index in real time. This is where most citations happen and where optimization actually moves the needle.
From the Ahrefs study of 1.4 million prompts (February 2025):
Search index sources get cited at 88.46%. When ChatGPT retrieves your page through Bing, it usually credits you. That's a high conversion rate from retrieval to citation.
Reddit gets retrieved constantly but cited at 1.93%. In fact, 67.8% of non-cited URLs in ChatGPT responses came from Reddit. The system uses it for context but almost never surfaces it as a credited source.
Natural language URL slugs get cited at 89.78% vs. 81.11% for opaque URLs. The difference is real and structural.
The practical implication: your biggest leverage point with ChatGPT is getting into Bing's index in a form the retrieval system recognizes as credible. That means clean URL structure, structured content, and content that answers specific questions directly, not general topic coverage. The formatting checklist we use for exactly that is in how to write content AI systems extract and cite.
How does Perplexity decide what to cite, and why is the pipeline different?
Answer capsule: Perplexity crawls the web on nearly every query, evaluates 60+ sources per response through a six-stage pipeline, and embeds citations into prompts before the language model generates anything. 70% of its top citations were updated within the last 12–18 months (ZipTie.dev, 2025–2026). 92.78% of pages it cites have fewer than 10 referring domains, meaning domain authority is not the primary signal.
Perplexity is architecturally different from ChatGPT in one critical way: it searches the web for nearly every query. ChatGPT selectively activates retrieval. Perplexity almost always retrieves.
This creates more citation opportunities for smaller sites. A new page indexed today can appear in Perplexity citations within hours or days. ChatGPT's search mode works similarly for recent content, but Perplexity's real-time crawling is more aggressive.
Three findings from ZipTie.dev's analysis of Perplexity's pipeline (2025–2026) matter most:
Freshness. 70% of Perplexity's top citations were updated within the past 12–18 months. A stale but comprehensive page loses citation spots to a thinner but recently refreshed page on time-sensitive queries. Publishing new content and updating existing content are both active citation signals.
Schema. Pages with schema markup achieve a 47% top-3 citation rate vs. 28% for pages without schema. FAQPage and Article schema are the most impactful types. They give Perplexity's pre-generation pipeline structured data to embed directly into prompts.
Domain authority is not the barrier. 92.78% of pages Perplexity cites have fewer than 10 referring domains. A well-structured page on a new domain can earn top-3 Perplexity citations against established competitors. This is the most important stat for any small business that has assumed AI citation requires years of SEO authority.
Perplexity also averages 21.87 citations per response, more than any other major AI platform (Qwairy, 118,101-answer study, Q3 2025). More slots means more opportunity.
What does the 11% source overlap between ChatGPT and Perplexity mean for your strategy?
Answer capsule: Only 11% of domains cited by ChatGPT overlap with domains cited by Perplexity, according to Averi's analysis of 680 million AI citations (March 2026). This means a strategy built for one platform leaves 89% of the other platform's citation landscape untouched. The content signals shared by both systems are the highest-leverage starting point.
An Averi analysis of 680 million AI citations (March 2026) found only 11% of domains are cited by both ChatGPT and Perplexity. A separate Passionfruit analysis confirmed a similar 12% overlap across three AI platforms.
The practical reading: if you're cited by ChatGPT, you are probably not cited by Perplexity. And vice versa. The platforms pull from different source pools, apply different ranking signals, and serve different citation patterns.
This is why tracking your visibility on one platform tells you almost nothing about the other. Businesses that see ChatGPT citations growing can still have zero Perplexity presence, and ChatGPT users and Perplexity users are different audiences with different query patterns.
The shared signals that work for both:
- First-hand, specific content that reads like operator knowledge
- Direct answers in the first 100 words of each section
- FAQPage schema
- Primary source citations with dates
- Named author with verifiable credentials
These are the E-E-A-T signals covered in detail here, and they are the highest-leverage starting point because they work across both systems rather than optimizing for one at the expense of the other.
For the platform-specific mechanics, what Perplexity cites and how to get in, or how Google AI Overviews fit into this picture, those are covered in the next posts in this series.
What content format gets cited by both ChatGPT and Perplexity?
Answer capsule: Content that gets cited by both ChatGPT and Perplexity shares five structural signals: a direct answer in the first 100 words, question-framed headings, FAQPage schema, named primary sources with dates, and first-hand framing that contains observations no AI could generate from training data. Schema alone lifts Perplexity top-3 citation rate from 28% to 47% (ZipTie.dev, 2025–2026).
This is not theoretical. On this blog, a post evaluating Square Issuing V2 used all five signals: first-hand evaluation framing, named primary sources (Federal filings, Marqeta documentation, Square's own pricing pages), question-phrased H2 headings, FAQPage schema, and direct answers in the opening section of each part.
That post generated 932+ Bing AI citations across query variants on a new domain with no established Google authority and minimal backlinks.
Citation volume started at zero and peaked at 171 per day on May 7, 2026, settling at 56–111 per day. During this period, Google organic clicks were near zero. AI citation volume preceded Google ranking authority by weeks.
The five signals in practice:
Answer in the first 100 words. 90% of Perplexity's top citations answer the section question within the first 100 words (ZipTie.dev). ChatGPT's citation selection also concentrates in early page content. The pattern is BLUF, Bottom Line Up Front. State the answer, then support it. Most web content does the reverse.
Question-framed headings. Both systems parse headings as question-answer pairs. A heading phrased as a question followed by a direct answer creates an extractable unit the system can embed into a prompt without surrounding context. For the mechanics of building these, see structuring content for AI extraction.
FAQPage schema. Perplexity citation rate jumps from 28% to 47% at top-3 with schema (ZipTie.dev). ChatGPT's source selection is also positively weighted by structured data. The FAQ section at the end of this post uses FAQPage schema, not for SEO decoration but because it creates extractable question-answer pairs for both systems.
Primary source links. Both ChatGPT and Perplexity favor content that itself points toward authoritative sources. A page that cites a government document, a named study, or a company's official documentation is treated as more credible than a page asserting the same information without sourcing.
First-hand framing. AI systems are trained on text. Content that contains observations a model could have generated from training data is interchangeable with that training data. Content that contains a specific client outcome, a documented pattern, or a before/after comparison that no model could fabricate is irreplaceable. That irreplaceability is the citation signal.
How do you verify that AI systems are citing your business?
Answer capsule: The primary tool for tracking AI citations is Bing's AI Performance dashboard, launched February 10, 2026, with Citation Share added June 16, 2026 (Microsoft Bing blog). It shows total citations, grounding queries (the phrases Copilot generates to retrieve your content), and page-level citation activity. For Perplexity and ChatGPT referrals, check GA4 for sessions from ai.perplexity.ai and chatgpt.com.
Most businesses have no idea whether AI systems are citing them. The tools to find out are newer than most guides acknowledge.
Bing AI Performance (the primary signal): Microsoft launched the AI Performance dashboard on February 10, 2026. It tracks citations across Microsoft Copilot, Bing AI summaries, and select partner integrations. The key metric: grounding queries, the internal phrases Copilot generates to retrieve your content. These are the actual questions your pages are being pulled to answer. You can see exactly which posts are being retrieved for which query types.
In June 2026, Microsoft added Citation Share, the first metric from a major platform that shows your percentage of citations within a topic, not just a raw count. As of June 16, 2026, it's live in preview worldwide.
GA4 referral attribution: Perplexity referrals appear in GA4 as sessions from ai.perplexity.ai. ChatGPT Search referrals appear from chatgpt.com. Neither platform sends high click volume relative to citation counts. The ratio of AI citations to referral clicks is much lower than traditional search. But even a small number of referrals confirms the system is citing and linking you.
Manual spot-checks: Search Perplexity directly for queries your business should answer. Check whether you appear. For ChatGPT, use web-search-enabled queries in the same category. A quick pass of 10–15 queries your business owns takes under 30 minutes and tells you immediately whether you're in the citation pool.
The tracking section is where most businesses discover they have a problem. If you're not in Bing Webmaster Tools, set it up before optimizing anything else. You can't improve what you're not measuring.
Frequently Asked Questions
Does ChatGPT cite small business websites, or only large publications?
ChatGPT cites small business websites when they appear in Bing's search index in a form the retrieval system recognizes as credible. The Ahrefs 1.4M prompt study (February 2025) found that search index sources get cited at 88.46%. When ChatGPT retrieves your page, it usually credits it. The barrier is getting retrieved in the first place, which depends on content structure and Bing indexing, not domain size.
How long does it take to start appearing in ChatGPT or Perplexity citations after publishing a new post?
Perplexity's real-time crawl means new content can appear in citations within hours to days of indexing. ChatGPT's search-enabled mode also operates on Bing's live index, so newly indexed pages can appear quickly. The variation is in how long it takes Bing to crawl and index the page, typically 24–72 hours for a site with an established crawl budget.
Is it possible to get cited by Perplexity without being cited by ChatGPT?
Yes, and it's common. Only 11% of domains are cited by both platforms (Averi, 680M citation study, March 2026). Perplexity's pipeline favors freshness, topical depth, and structured content regardless of domain authority. ChatGPT's selection favors established editorial sources and Bing-indexed content with clean URL structure. You can earn strong Perplexity visibility before ChatGPT ever cites you.
Do AI systems cite the same page over and over, or do they rotate sources?
They rotate. Query wording changes citation results. Perplexity assembles fresh source sets on each query via real-time retrieval. ChatGPT Search uses live Bing results, which vary by query phrasing and timing. A page cited for "merchant account fees" may not be cited for "payment processing fees for small business," which is why citation count totals, not just single-query checks, matter for measurement.
What is a "grounding query" in Bing Webmaster Tools?
A grounding query is the internal phrase Microsoft Copilot generates when it needs to retrieve web content to answer a user's question. These are not the exact words the user typed. They are reformulated queries the system uses to find relevant sources. Bing Webmaster Tools' AI Performance dashboard makes these visible, which lets you see what questions your content is being pulled to answer versus what questions you expected to own. The gap between those two lists is usually where the content work is.
Should I optimize for ChatGPT and Perplexity separately, or use the same content for both?
Use the same foundational content. The signals that work for both (BLUF answers, schema, primary sources, first-hand framing) are the right starting point. Then layer platform-specific signals: for Perplexity, prioritize freshness and topical depth over domain authority; for ChatGPT, focus on Bing indexing, URL structure, and presence in authoritative editorial sources. The same well-structured post, kept current and properly marked up, earns citations on both platforms. A post optimized only for one typically underperforms on the other.
Most guides on getting cited by AI tell you to add structured data and write clearly. That is the floor, not the ceiling.
The ceiling is content that contains what no AI can fabricate: specific, documented, first-hand observations from real work. That is what generated 932+ citations on a new domain, 171 per day at peak, AI visibility that arrived before a single Google organic click.
The method is the same whether we run it on this blog or yours.
Kaleb Dickhaut — Founder, ClickWerxs. Kaleb built ClickWerxs from the ground up — from payment processing ISO to the Command Center platform to the AI SEO methodology the blog runs on. He has onboarded hundreds of small businesses onto payment and CRM systems. linkedin.com/in/kaleb-dickhaut
Sources
- StatCounter Global Stats, search engine market share worldwide — Google 91.25% as of June 2026. gs.statcounter.com
- ClickWerxs blog, first-party AI citation data, May 2026 — one post generated 932+ Bing AI citations across query variants from a new domain; daily citations rose from 0 to a peak of 171 on 7 May 2026, settling at 56–111 per day, with Google organic clicks near zero over the same period. Reported as a past result, not a promise of performance. Operator data.
ClickWerxs sells SEO and AI visibility services and earns revenue from those engagements. First-party figures are past results for this blog and are not a promise of future performance. This is operator opinion, not professional advice.