Community platforms — LinkedIn, Reddit, Medium, Substack, GitHub, and Y Combinator's Hacker News — account for just 4.1% of all AI search citations tracked by the Machine Relations Index, but they produce the widest engine-by-engine divergence of any source category. Google AI Mode generates 234 community-platform citations; ChatGPT generates 4. That 58x gap makes community content the single most engine-dependent variable in AI visibility strategy.
Methodology #
This analysis draws from the Machine Relations Index (MRI v2), which measures source-segment citation rates across six AI engines — Perplexity, ChatGPT, Gemini, Claude, Google AI Mode, and Google AI Overviews — using 17,540 citation events from 6,020 domains. Citation rates are published only for segments that clear the evidence floor of at least 10 observations across at least 7 distinct run dates. Community and social platforms are classified under the MRI source-role taxonomy as domains where users publish content directly (forums, professional networks, blogging platforms, code repositories) rather than through an editorial or vendor-controlled process. The MRI tracks 20 domains in this category. MRI methodology details.
The Community Citation Landscape #
Twenty community platforms generated 712 total citations, yielding a per-domain density of 35.6 citations — the highest of any MRI source category. For comparison, the 4,158 domains in the "uncategorized" long-tail category average just 1.8 citations each, and even editorial media (545 domains, 2,594 citations) averages only 4.8.
Three platforms dominate: LinkedIn (251 citations), Reddit (189), and Medium (139). Together they account for 81% of all community-platform citations. The remaining 17 platforms — including Y Combinator (40), Substack (39), GitHub (17), dev.to (8), and Facebook (8) — contribute the other 19%.
Community Platform Citation Rankings #
| Platform | Total Citations | Engines Citing | Subject Categories | Confidence |
|---|---|---|---|---|
| linkedin.com | 251 | 5 of 6 | 10 | A |
| reddit.com | 189 | 4 of 6 | 10 | A |
| medium.com | 139 | 5 of 6 | 10 | A |
| ycombinator.com | 40 | 5 of 6 | 6 | B |
| substack.com | 39 | 5 of 6 | 8 | B |
| github.com | 17 | 4 of 6 | 4 | B |
| dev.to | 8 | 4 of 6 | 2 | C |
| facebook.com | 8 | 2 of 6 | 5 | C |
Source: Machine Relations Index v2, measurement window ending July 2026. 17,540 total citation events across 6,020 domains. Confidence grades reflect evidence volume: A = high evidence, B = moderate, C = limited.
Engine-by-Engine Community Preferences #
The defining feature of community-platform citations is how differently each engine treats them. No other source category produces this degree of variation.
Per-Engine Community Citation Breakdown #
| Engine | Community Citations | % of Engine Total | Top Community Source |
|---|---|---|---|
| Google AI Mode | 234 | 5.5% | LinkedIn (123) |
| Perplexity | 179 | 6.2% | Reddit (75) |
| Google AI Overviews | 154 | 7.1% | LinkedIn (62) |
| Gemini | 119 | 2.5% | Medium (65) |
| Claude | 22 | 1.4% | Medium (12) |
| ChatGPT | 4 | 0.2% | Medium (1) |
This data aligns with and extends several independent studies. SurfacedBy's analysis of 127,198 citations found that 69.6% of cited domains appear in only one engine's results — a pattern most pronounced in community content. TryProfound's platform comparison found Perplexity "prioritizes community discussions" while ChatGPT emphasizes "authoritative knowledge bases." Conductor's seven-month tracking study confirmed that Google AI Mode "uniquely cited LinkedIn for Education intent — no other engine did this," supporting the MRI's finding that LinkedIn is Google AI Mode's dominant community source.
What Each Engine Prefers #
Google AI Mode is the heaviest community citer at 5.5% of its total output. LinkedIn alone accounts for 123 of its 234 community citations — making LinkedIn Google AI Mode's single most-cited community domain and its fifth-most-cited domain overall. Google AI Mode also draws from Reddit (44), Medium (42), and Substack (11). This is consistent with Search Engine Journal's finding that Google's AI surfaces embrace user-generated content at rates of 18% or higher. Frase's cross-engine analysis similarly identified Google AI Mode as an outlier in its willingness to cite non-traditional web sources.
Perplexity concentrates community citations on Reddit (75 of 179), making Reddit its single most-cited domain across all source categories. LinkedIn follows at 64 and Y Combinator at 15. Perplexity does not cite Medium. This reflects Perplexity's well-documented Reddit affinity — it treats Reddit threads as real-time market evidence.
Google AI Overviews distributes community citations more evenly: LinkedIn (62), Reddit (54), Medium (19), with smaller contributions from Y Combinator (8) and Facebook (5). AI Overviews generates 7.1% of its citations from community platforms — the highest rate of any engine — but no single platform dominates as heavily as LinkedIn does for AI Mode or Reddit does for Perplexity.
Gemini concentrates on Medium (65 of 119 community citations), making Medium Gemini's preferred community source by a wide margin. Reddit receives 16 citations from Gemini; LinkedIn receives just 1. Gemini's Medium preference is notable because Medium content often mirrors the technical depth and formatting conventions of editorial publications.
Claude generates just 22 community citations (1.4% of its total). Medium accounts for 12, Y Combinator for 4. Claude cites neither LinkedIn nor Reddit. Claude's documented preference for market databases (13.7% of its citations vs. 6–9% for other engines) suggests it deprioritizes informal or discussion-format content. Aether AI's comparative analysis of engine citation differences attributed this to Claude's emphasis on "verified, structured data sources" over discussion-format content.
ChatGPT is effectively absent from community citations: 4 total across all 20 platforms. One from Medium, two from GitHub, one from a niche forum. This is the starkest finding in the dataset. ChatGPT's top-cited sources are editorial publications — TechRadar (131), Axios (63), ITPro (51) — and it allocates 29% of all citations to editorial media versus 11–13% for other engines. Community content simply does not enter its citation pipeline. Tinuiti's 2026 citation tracking noted that ChatGPT citation patterns "differ fundamentally from other AI search surfaces," and the MRI data confirms this divergence is most extreme for community-sourced content.
LinkedIn: The Most-Cited Community Platform #
LinkedIn's 251 total citations place it among the top 10 most-cited domains across the entire MRI — higher than many traditional editorial publications. But the distribution is extreme:
| Engine | LinkedIn Citations |
|---|---|
| Google AI Mode | 123 |
| Perplexity | 64 |
| Google AI Overviews | 62 |
| Gemini | 1 |
| Claude | 1 |
| ChatGPT | 0 |
Google's two AI surfaces account for 73.7% of all LinkedIn citations. This means LinkedIn visibility is almost entirely a Google AI play. Brands investing in LinkedIn content for AI search visibility should expect returns from Google's AI products but little impact on ChatGPT, Claude, or Gemini citations.
LinkedIn earns the highest confidence grade (A) among community platforms, backed by citations from 5 of 6 engines across all 10 measured subject categories. This breadth indicates universal topical relevance even though citation volume concentrates in Google's products.
Reddit: Perplexity's Primary Source #
Reddit is the only community platform where a single engine accounts for the majority of its non-Google citations. Perplexity generates 75 of Reddit's 189 total citations (39.7%), making Reddit Perplexity's most-cited domain across all categories.
The Reddit citation pattern across engines reveals a clear split between search-adjacent engines (Perplexity, Google AI Mode, Google AI Overviews) and model-native engines (ChatGPT, Claude):
- Search-adjacent engines cite Reddit 173 times combined
- Model-native engines cite Reddit 0 times
This divide suggests that engines with active web-retrieval architectures — which crawl and index live content — naturally encounter Reddit's high-frequency discussion threads. Engines that rely more heavily on parametric knowledge or curated retrieval pipelines bypass community discussions entirely. Digital Applied's research found that 18% of AI citations come from sources outside the traditional top 100 search results, suggesting that AI engines actively seek non-traditional sources like Reddit discussions that rank poorly in conventional search but carry high informational value.
Medium: Gemini's Community Preference #
Medium's 139 citations make it the third-most-cited community platform, but its engine distribution is unlike any other. Gemini generates 65 of Medium's citations (46.8%), more than double the next engine (Google AI Mode at 42).
Medium's elevated Gemini presence may reflect formatting: Medium articles tend to follow editorial conventions (clear headings, structured arguments, bylined authorship) that align with the signals Gemini uses to assess source credibility. SurfacedBy's research found that Gemini averages 11.0 sources per answer — the highest of any engine — suggesting it casts a wider retrieval net that captures well-structured community content. Campaign Creators' citation analysis noted that "AI citation patterns align more closely with text-based authority signals, particularly multi-platform brand mentions," which may explain why Medium's author-branded article format earns Gemini citations that less-structured community platforms do not.
The Emerging Tier: Substack, Y Combinator, and GitHub #
Below the top three, a second tier of community platforms shows distinct citation patterns:
Substack (39 citations, confidence B) is cited by 5 of 6 engines — the broadest reach of any community platform outside the top three. Its citations distribute relatively evenly: Gemini (14), Google AI Mode (11), Perplexity (8), Google AI Overviews (4), Claude (2). Substack's newsletter format and subject-matter expertise may contribute to its cross-engine acceptance.
Y Combinator (40 citations, confidence B) draws from the startup and technology community. Perplexity leads (15), followed by Gemini (10), Google AI Overviews (8), Claude (4), and Google AI Mode (3). Hacker News discussions appear as cited sources in enterprise AI, fintech, and infrastructure queries.
GitHub (17 citations, confidence B) represents the only code-repository platform with measurable AI citations. Gemini (7), Perplexity (5), Google AI Mode (3), and ChatGPT (2) cite it. GitHub's citation presence suggests that AI engines increasingly treat code repositories and technical documentation as citation-worthy sources alongside traditional publications.
Implications for AI Visibility Strategy #
The community-platform data creates three actionable findings for brands managing Machine Relations across multiple engines:
1. Community investment is engine-specific. LinkedIn content improves Google AI citations. Reddit participation improves Perplexity citations. Medium publishing improves Gemini citations. None of these investments move ChatGPT or Claude citations. Brands that treat "community content" as a single channel will misallocate effort.
2. Per-domain citation density dwarfs other categories. Community platforms average 35.6 citations per domain versus 4.8 for editorial media and 1.8 for the uncategorized long tail. Each community platform with confidence A or B evidence delivers citation volume comparable to dozens of smaller editorial placements.
3. ChatGPT's community blindspot is structural, not accidental. With only 4 community citations from 1,897 total, ChatGPT's avoidance of community content appears to be a retrieval-pipeline characteristic, not a data gap. ChatGPT allocates 29% of citations to editorial media — the highest of any engine — suggesting it actively filters for editorially-controlled sources. GeoXylia's cross-engine comparison arrived at a similar conclusion, noting that ChatGPT and Claude "demonstrate a preference for institutionally-published content" while Google's surfaces favor source diversity. Brands dependent on ChatGPT visibility should not expect community strategies to contribute.
FAQ #
Do AI engines cite Reddit? Four of six engines cite Reddit, but concentration varies dramatically. Perplexity generates 75 Reddit citations (the most of any engine), Google AI Overviews generates 54, and Google AI Mode generates 44. ChatGPT and Claude generate zero Reddit citations in the MRI measurement window.
Does LinkedIn content help with AI search? LinkedIn is the most-cited community platform (251 total citations) and carries the highest MRI confidence grade (A) in this category. However, 73.7% of LinkedIn citations come from Google's two AI surfaces (AI Mode and AI Overviews). LinkedIn content has minimal impact on ChatGPT, Claude, or Gemini AI citations.
Which AI engine cites the most community content? Google AI Overviews allocates the highest percentage of its citations to community platforms (7.1%), while Google AI Mode generates the highest absolute count (234 citations). ChatGPT allocates the least (0.2%, just 4 citations total).
Should brands invest in Medium for AI visibility? Medium is an effective channel specifically for Gemini visibility. Gemini generates 65 of Medium's 139 total citations, more than any other engine. Medium also receives meaningful citations from Google AI Mode (42) and Google AI Overviews (19).
Does Substack content get cited by AI engines? Substack achieves the broadest cross-engine reach of any mid-tier community platform, cited by 5 of 6 engines with 39 total citations. Its newsletter format appears to earn citations from Gemini, Google AI Mode, Perplexity, Google AI Overviews, and Claude.
Last updated: July 21, 2026. Data: Machine Relations Index, 30-day window, 17,540 citation events, 6,020 domains, 6 engines. MRI methodology.