What Types of Content Do AI Search Engines Cite?
Across 1,629 classified citations, first-party vendor pages were the largest source type for all five engines. ChatGPT drew 67.8% of its classified citations from first-party sites, while Perplexity drew 39.4% and relied more heavily on business press and review platforms.
Methodology
The classified subset of the Q4 2026 Citation Benchmark contains 1,629 citations from 375 buyer-intent responses across 75 prompts × 5 engines. Foglift sent the same frozen prompt set to ChatGPT, Claude, Gemini, Google AI Overview, and Perplexity on October 2, 2026. Cited URLs were normalized to domains and counted once per response, then mapped to one of fifteen structural categories defined in state/research/domain-taxonomy.json. The taxonomy covers vendor first-party sites, niche publisher hubs, listicles, business press, tech press, lifestyle media, institutions, video, community discussion, review platforms, marketplaces, personal blogs, social media, developer platforms, and search-engine cache URLs. High-volume domains were hand-classified; the long tail remains unclassified rather than being assigned from weak evidence. We report 1,629 classified citations, or 51.3% of the 3,173 domain citations. Every per-engine percentage uses that engine's classified subset as its denominator. Provider models and indexes changed between Q2 and Q4, so quarter-over-quarter differences are observational, not causal.
First-party pages lead, but every engine has a different source mix
The Q4 benchmark measured the cross-engine citation graph at the domain level. This report asks a different question: what kind of source does each engine cite? We classified 1,629 citations into fifteen structural categories, then calculated each category as a share of the engine's classified citations.
Vendor first-party sites are the largest category for every engine. The share ranges from 67.8% for ChatGPT to 39.4% for Perplexity. Beyond that shared anchor, the mix separates quickly. Google AI Overview and Gemini carry the largest video and community shares. Perplexity carries 67.1% of all business-press citations and 78.7% of all review-platform citations in the classified sample. A single universal distribution plan would miss those engine-level differences.
Population (n = 5 engines × 75 buyer-intent prompts)
375
Responses sampled
3173
Citations extracted
1629 (51.3%)
Citations classified
15
Categories represented
25
Verticals covered
Q4 2026 benchmark
Source dataset
The per-engine summary table
Each engine returned 75 responses, but the count of citations per response varies by almost fivefold. Perplexity leads at 16.24 citations per response, followed by Gemini at 9.36 and Google AI Overview at 8.63. Claude returned 4.71 and ChatGPT 3.37. The classified subset within each engine is the denominator for the structural comparison.
| Engine | Citations / response | Classified | Classified % | Top category |
|---|---|---|---|---|
| ChatGPT | 3.37 | 152 / 253 | 60.1% | Vendor first-party (67.8%) |
| Claude | 4.71 | 154 / 353 | 43.6% | Vendor first-party (49.4%) |
| Gemini | 9.36 | 370 / 702 | 52.7% | Vendor first-party (51.1%) |
| Google AI Overview | 8.63 | 321 / 647 | 49.6% | Vendor first-party (47.0%) |
| Perplexity | 16.24 | 632 / 1218 | 51.9% | Vendor first-party (39.4%) |
The full per-engine, per-category matrix
Each cell is the share of the engine's classified citations going to that category. The categories are ordered by total classified citation volume across all five engines. A 0% cell means the Q4 sample contained no classified citation from that engine to that category. It does not establish that the engine never cites that source type.
| Content type | ChatGPT | Claude | Gemini | AIO | Perplexity |
|---|---|---|---|---|---|
| Vendor first-party | 67.8% | 49.4% | 51.1% | 47.0% | 39.4% |
| Niche publisher hub | 13.2% | 21.4% | 14.6% | 15.6% | 18.8% |
| Business press | 3.3% | 7.1% | 4.6% | 4.4% | 15.2% |
| Listicle / content-farm | 2.0% | 10.4% | 5.1% | 5.6% | 6.5% |
| Video (mostly YouTube) | 0.7% | 0.0% | 9.5% | 11.2% | 0.0% |
| Tech press | 2.0% | 0.0% | 2.4% | 2.8% | 4.3% |
| Review aggregator | 1.3% | 1.3% | 1.1% | 0.6% | 5.9% |
| Community discussion (Reddit, Quora, HN) | 0.7% | 0.0% | 5.1% | 6.5% | 0.3% |
| Institutional (.gov / .edu / nonprofit) | 3.3% | 3.9% | 2.4% | 1.2% | 2.2% |
| Marketplace | 2.6% | 1.9% | 1.4% | 1.6% | 2.5% |
| Search-engine cache URL | 1.3% | 3.2% | 0.3% | 0.6% | 1.6% |
| Lifestyle media | 1.3% | 0.0% | 0.3% | 0.3% | 2.5% |
| Personal blog / Medium | 0.0% | 0.6% | 1.1% | 1.9% | 0.3% |
| Social media | 0.0% | 0.6% | 0.5% | 0.6% | 0.2% |
| Developer platform | 0.7% | 0.0% | 0.5% | 0.0% | 0.3% |
The source mix moved between Q2 and Q4
The same 75-question prompt set produced a different source mix in Q4. ChatGPT's first-party share was nearly unchanged at 68.0% to 67.8%. The largest shifts appeared elsewhere: Google AI Overview's video share rose from 2.6% to 11.2%, Gemini's rose from 3.0% to 9.5%, and Perplexity's video share moved from 9.7% to 0%. Perplexity instead increased its business-press share from 1.6% to 15.2% and its review-platform share from 1.3% to 5.9%.
These are comparable snapshots, not a controlled attribution study. Provider models and retrieval indexes changed between collections. The practical conclusion is narrower: use current engine-level evidence, because a distribution assumption that fit one quarter may be wrong in the next.
| Engine | First-party | Video | Community | Reviews | Business press |
|---|---|---|---|---|---|
| ChatGPT | 68.0% → 67.8% | 0.0% → 0.7% | 0.0% → 0.7% | 0.0% → 1.3% | 2.7% → 3.3% |
| Claude | 52.2% → 49.4% | 0.0% → 0.0% | 0.0% → 0.0% | 0.5% → 1.3% | 3.9% → 7.1% |
| Gemini | 45.7% → 51.1% | 3.0% → 9.5% | 2.4% → 5.1% | 1.6% → 1.1% | 5.4% → 4.6% |
| Google AI Overview | 48.8% → 47.0% | 2.6% → 11.2% | 1.0% → 6.5% | 2.1% → 0.6% | 5.2% → 4.4% |
| Perplexity | 47.8% → 39.4% | 9.7% → 0.0% | 0.0% → 0.3% | 1.3% → 5.9% | 1.6% → 15.2% |
Coverage: how broad each engine's citation diet is
The fifteen classified categories are not equally represented across engines. Gemini cites all fifteen. Google AI Overview covers fourteen, missing only developer platforms. Perplexity covers fourteen, missing only video. ChatGPT covers thirteen, while Claude covers ten and has no classified video, tech-press, community, lifestyle-media, or developer-platform citation in this sample.
| Engine | Categories cited | Categories absent |
|---|---|---|
| ChatGPT | 13 / 15 | Personal blog / Medium, Social media |
| Claude | 10 / 15 | Video (mostly YouTube), Tech press, Community discussion (Reddit, Quora, HN), Lifestyle media, Developer platform |
| Gemini | 15 / 15 | (all covered) |
| Google AI Overview | 14 / 15 | Developer platform |
| Perplexity | 14 / 15 | Video (mostly YouTube) |
Five engine signatures
Run the per-engine columns through the eye and a personality emerges for each one. Below is the simplest version of that read.
ChatGPT
The strongest first-party concentration. Vendor sites account for 67.8% of its classified citations. Video, community, and review platforms each remain below 1.5%.
Top vendor first-party citations
- hubspot.com (3)
- activecampaign.com (3)
- zillow.com (3)
- mailchimp.com (2)
- microsoft.com (2)
Claude
A first-party and niche-publisher mix. Vendor sites account for 49.4% and niche publisher hubs 28.6%. The sample contains no video or community citation.
Top niche publisher hubs
- nortonmediaenterprise.com (3)
- outdoorgearlab.com (3)
- mattressclarity.com (3)
- sleepfoundation.org (3)
- upgradedpoints.com (2)
Gemini
The broadest mix. Gemini cites all fifteen categories, with 51.1% first-party, 9.5% video, and 5.1% community discussion.
Top community UGC citations
- reddit.com (19)
Google AI Overview
The strongest video and community mix. Video accounts for 11.2% of its classified citations and community discussion accounts for 6.5%.
Top video citations
- youtube.com (36)
Perplexity
The press and review outlier. Business press accounts for 15.2% and review platforms 5.9%, while no video citation appears in its classified Q4 sample.
Top business-press citations
- forbes.com (30)
- nytimes.com (12)
- indiatimes.com (8)
- cnn.com (8)
- businessinsider.com (8)
The vertical overlay: SaaS is vendor-first, CPG is publisher-first
Collapse the 25 verticals into three super-categories: tech SaaS, consumer services, and CPG / retail. The engine-level patterns above are real, but they are partly an artifact of which verticals each prompt set covered. Looking at the vertical lens directly, the gap between SaaS and CPG / retail is much larger than the gap between any two engines.
| Content type | Tech SaaS | Consumer services | CPG / retail |
|---|---|---|---|
| Vendor first-party | 76.1% | 48.7% | 10.7% |
| Niche publisher hub | 0.5% | 17.0% | 37.1% |
| Business press | 2.5% | 6.5% | 17.5% |
| Listicle / content-farm | 6.7% | 4.0% | 5.9% |
| Video (mostly YouTube) | 2.5% | 2.9% | 7.4% |
| Tech press | 2.4% | 5.8% | 2.3% |
| Review aggregator | 5.9% | 1.1% | 0.0% |
| Community discussion (Reddit, Quora, HN) | 1.3% | 4.7% | 3.3% |
| Institutional (.gov / .edu / nonprofit) | 0.0% | 4.3% | 4.3% |
| Marketplace | 0.3% | 0.4% | 5.0% |
| Search-engine cache URL | 0.1% | 2.5% | 2.0% |
| Lifestyle media | 0.0% | 1.1% | 2.8% |
| Personal blog / Medium | 0.5% | 1.1% | 1.0% |
| Social media | 0.3% | 0.0% | 0.7% |
| Developer platform | 0.7% | 0.0% | 0.0% |
Tech SaaS responses cite the vendor's own first-party 76.1% of the time and a niche publisher hub 0.5% of the time. CPG / retail responses cite the vendor first-party 10.7% of the time and a niche publisher hub 37.1% of the time. The vertical changes which source layer is the most plausible next test. A SaaS team can start with a precise product or developer page. A consumer brand also needs category publishers, press, video, marketplaces, or community evidence in the plan.
Three categories concentrate in Perplexity
Three categories send at least 60% of their Q4 citations to one engine:
- Business press: Perplexity accounts for 67.1% of all 143 citations to this category in the dataset.
- Review aggregator: Perplexity accounts for 78.7% of all 47 citations to this category in the dataset.
- Lifestyle media: Perplexity accounts for 80.0% of all 20 citations to this category in the dataset.
Perplexity accounts for 96 of 143 business-press citations, 37 of 47 review-platform citations, and 16 of 20 lifestyle-media citations. Those source types are therefore more relevant to a Perplexity test than to a universal engine plan. The panel does not establish that earning one of those placements will cause a citation.
What this means for a publisher
This is an observational source-share panel. It records which source types appeared in one frozen set of answers. It does not identify why an engine selected a source or estimate the citation lift caused by publishing, editing, or earning a mention. The actions below are channel hypotheses to test with stable prompts and engine-level before-and-after measurement.
One. Test vendor first-party surfaces first. Every engine puts the brand's own site first in the Q4 classified sample (ChatGPT 67.8%, Gemini 51.1%, Claude 49.4%, Google AI Overview 47.0%, Perplexity 39.4%). That makes product pages, resource centers, glossary entries, and FAQs the clearest owned-surface hypothesis in this sample. An intervention test is still required. Improve the page that answers the tracked prompt, then rerun the same prompt by engine. The related AI Readiness study found that 23.2% of scanned sites ship no structured data at all, which is a separate technical observation rather than evidence of citation lift.
Two. Test niche-publisher-hub placement for consumer-facing categories. In CPG, retail, and health prompts, the observed mix includes category publishers such as Sleep Foundation, Healthline, OutdoorGearLab, and The Good Trade. Across all engines combined, niche publisher hubs are the second-largest classified content type after vendor first-party. That makes an independent placement test reasonable; it does not prove that a placement will produce a citation.
Three. Test video and community discussion by engine, not as universal channels. In this sample, video was concentrated in Google AI Overview and Gemini, with a small ChatGPT share and none on Claude or Perplexity. Community citations also concentrated in Google AI Overview and Gemini. Use those distributions to choose a test, then measure the target engines independently instead of assuming the sample predicts future behavior.
How to use this source-layer map
The dataset is useful because it separates the owned-page work from the outside-source work. When an engine's observed mix is mostly first-party vendor pages, the next test can start on product, docs, FAQ, comparison, and research pages. When the observed mix includes publisher hubs, community threads, reviews, or video, test the matching outside-source channel separately.
Use the table above to pick the channel by engine, then use the playbooks below to turn that channel into a concrete action queue.
Engine fit: Test when the target engine's observed mix includes business press, tech press, or niche publisher hubs.
Action: Package original data with sample size, methodology, and a short quote journalists can cite.
Engine fit: Test when first-party pages are prominent in the target engine's observed mix.
Action: Build a complete owned cluster with glossary, comparison, FAQ, product, and research pages.
Engine fit: Test on prompts where the answer calls for a named expert or repeatable category claim.
Action: Turn one defensible claim into first-party copy, guest commentary, podcast notes, and citations.
Engine fit: Test primarily for Google AI Overview and Gemini, which supplied 40 of 43 community citations in this classified sample.
Action: Answer real buyer questions in public communities, then reinforce the same language on owned pages.
Engine fit: Test where the target engine's observed mix includes review aggregators, community forums, or profile pages.
Action: Keep review profiles accurate, structured, and aligned with the factual claims on your site.
Video with a crawlable transcript
Engine fit: Test primarily for Google AI Overview and Gemini, which had 11.2% and 9.5% video shares in the Q4 classified sample.
Action: Publish a crawlable walkthrough with transcript, matching page links, and clear product/entity language.
Other reports in this series
This Q4 refresh uses the same frozen 75-question prompt set as the quarterly citation benchmark. The related reports below add the domain, cross-engine, buyer-intent, and publisher-readiness views.
- AI engines agree on brands more than sources. A post-fix production-data study of answer agreement vs. source overlap.
- AI Readiness across 194 websites. Median AI Readiness 60 vs median SEO 90. The 30-point publisher-side gap.
- Top 100 most-cited domains in AI search (Q2 2026). Only 12 of 1,119 domains are cited by all five engines.
- ChatGPT vs. Google AI Overview: the same prompt, two different webs. 3.4% average Jaccard overlap, with no shared cited domain on 78.7% of paired prompts.
- Buyer intent reshapes AI citations. Discovery, shortlist, and variation cite different domain sets.
- When AI engines cite the reviewer vs. the brand. The 70-point vendor-vs-aggregator gap by vertical.
- Top 100 most-cited domains in AI search, Q4 2026. The underlying dataset.
To turn the content-type mix into a monitoring plan, compare tool coverage in the AI search visibility comparison hub and track engine-specific prompt behavior with Foglift AI search monitoring.
Frequently Asked Questions
What types of content do AI search engines cite?
In Foglift's Q4 2026 classified sample, vendor first-party pages were the largest content type for all five engines and accounted for 47.1% of the classified citations overall. Niche publisher hubs ranked second at 16.9%. Press, video, community discussion, review platforms, marketplaces, and developer sources varied sharply by engine and vertical.
Why is the unclassified portion 48.7% of all citations?
The 15-category taxonomy was applied to high-volume domains. Long-tail domains, many cited once or twice, were left unclassified instead of being assigned from weak evidence. Every per-engine percentage uses the classified subset as its denominator. Read the percentages as the distribution within the classified high-volume subset, not as a complete census of every citation.
ChatGPT sends 67.8% of its classified citations to vendor first-party sites. Does that make it the easiest engine to win?
No. The 67.8% result describes where citations landed within ChatGPT's classified subset. It does not show that editing first-party pages will cause more citations or that ChatGPT is easier to influence. It supports a testable channel choice: improve the product, documentation, glossary, FAQ, or research page that directly answers the prompt, keep the prompt stable, and measure the outcome by engine.
Which engines cited video in Q4?
Google AI Overview had the largest video share at 11.2% of its classified citations, followed by Gemini at 9.5%. ChatGPT was 0.7%, while Claude and Perplexity were 0%. This is a sharp change from the Q2 sample, where Perplexity had the largest video share. The study observes returned citations and does not infer a provider mechanism.
Which engines cited community discussion in Q4?
Google AI Overview accounted for 21 of the 43 community citations, Gemini accounted for 19, Perplexity returned two, ChatGPT returned one, and Claude returned none. This panel measures cited URLs, not model training or the downstream effect of community participation. Treat community work as an engine-specific test and measure the same prompts before and after it.
How does this connect to the Q4 2026 benchmark?
The Q4 benchmark ranks cited domains across the frozen 75-question panel. This report projects the same citations through a 15-category taxonomy and compares the source mix by engine and vertical. The companion Q2 report provides a prior snapshot, but provider models and search indexes changed between collections, so the differences should be treated as observed drift rather than a controlled causal result.
Free tool
Run a Technical Audit on your own site
Check technical SEO, performance, security, accessibility, and structural AI Readiness. No signup required.
Run Free Technical Audit