Forums and Reviews in AI Answers: How Much User Content Each AI Engine Cites

How much an AI engine cites forums and reviews depends mostly on whether it is a Google surface: about 17.5% of AI Overviews citations in one 2026 study, under 2% on ChatGPT, Perplexity and Gemini. Here is the engine-by-engine picture.

R
Rankry Team
· 12 min read · Updated

How much an AI engine relies on forums and reviews depends mostly on whether it is a Google surface. In BrightEdge’s April 2026 study, user-generated content made up about 17.5% of Google AI Overviews citations and about 7% of AI Mode’s, against 1.5% for Perplexity, 0.5% for ChatGPT and 0.2% for Gemini. OtterlyAI’s September 2026 data points the same way: forums and video took about a quarter of citations on both Google surfaces and under 3% on every other engine, and Claude did not cite reddit.com on any of 77 measured days. ChatGPT’s Reddit share fell from 3.765% (July 1 to August 7, 2026) to 0.227% in September after a drop that began August 8. Review platforms appear most often when the query asks for reviews: SE Ranking found at least one in 49% of AI Overviews for explicit review searches. So put community and review effort where your buyers’ engines actually read it, and do it honestly, because the FTC’s fake-review rule has been in force since October 21, 2024.

Most advice on Reddit and reviews treats “AI” as one audience. The data says there are at least three, and each study counts differently, so we present them one at a time instead of forcing them into one table.

Google’s AI surfaces read the most user content

Share of citations from user-generated content by AI engine, BrightEdge, April 2026

BrightEdge’s April 24, 2026 study compared five engines with a domain-level taxonomy. Its UGC bucket covers video platforms, forums, community sites and social networks; review sites sit in a separate commercial bucket. BrightEdge says its prompts span ten industries (it names nine) and does not disclose its sample size or collection period, so treat the figures as directional. UGC took about 17.5% of AI Overviews citations (18% in its table), about 7% in AI Mode, 1.5% in Perplexity, 0.5% in ChatGPT and 0.2% in Gemini. One unnamed video platform alone held about 10.6% of AI Overviews citations, and one unnamed forum another 2.9%.

OtterlyAI’s October 14, 2026 study is more recent and discloses its sample: 617,985 citations across 85,330 URLs, five industries, seven engines, US, September 1 to 30, 2026. In both AI Overviews and AI Mode, forums took 11% to 14% of citations and video another 12% to 14%. Reddit and YouTube supplied 92% of those citations in AI Overviews and 98% in AI Mode. On every other engine, forums and video stayed below 3%. One caveat: about three quarters of Otterly’s prompts ask for rankings, which favors listicles.

SE Ranking’s January 2026 study of 22,729 AI Overviews from 30,000 US commercial keywords, checked once on December 1, 2025, found social and community platforms (Reddit, YouTube, Quora, Medium, LinkedIn, Stack Exchange and similar) in 60.1% of AI Overviews and in 13.4% of more than 211,000 links.

The AI Mode numbers disagree: about 7% in BrightEdge’s April data, roughly a quarter in Otterly’s September data. Different prompts, dates and classifications likely explain part of it, and AI Mode itself moved fast: Otterly measured its YouTube share falling from 13.521% on September 1 to 3.877% on September 20. Read these as ranges, not a trend line. Why the two Google surfaces still cite different pages is covered in our overlap analysis of AI Overviews, AI Mode and ChatGPT sources.

Reddit, engine by engine

Reddit share of each engine's citations on September 20, 2026, OtterlyAI

One dataset that measures Reddit on seven engines with one method is OtterlyAI’s September 22, 2026 update: 1,600 prompts across 16 US industry reports, run daily, reported as the median of daily shares. On September 20, Reddit held 5.433% of AI Overviews citations, 5.264% of AI Mode, 2.043% of Perplexity, 1.295% of Gemini and 0.237% of ChatGPT. Claude did not cite reddit.com on any of the 77 days measured. Microsoft Copilot cited it on 23 of 77 days and never above 0.05%.

ChatGPT used to be different. Reddit held 3.765% of its citations from July 1 to August 7, 2026, then 0.553% from August 8 to 24, and 0.227% across September. An earlier Otterly study, published August 27 and based on 16 brand reports, found ChatGPT’s daily Reddit citations fell from 497 to 132 (-73.4%) while ChatGPT’s total citation volume rose 3.5% and Reddit citations on other engines held flat (-0.6%). Official brand and competitor pages gained in 10 of 15 reports. Otterly says its data does not show the cause; a change in how ChatGPT retrieves sources is its “leading hypothesis, not a conclusion.” OpenAI had not published an explanation as of that report. The full September picture is in September’s AI citation shift.

It was not the first cut. In a Semrush study of about 230,000 prompts tracked from July to October 2025, Reddit appeared in close to 60% of ChatGPT responses in early August 2025 and around 10% by mid-September. That metric is share of responses, not share of citations, so the two drops cannot be compared in size, only in direction.

For a longer baseline, Profound’s analysis of 680 million citations from August 2024 to June 2025 (its overall citation volume tables) put Reddit at 1.8% of ChatGPT citations, 2.2% of AI Overviews and 6.6% of Perplexity. Different method again, so do not read the gap to Otterly’s 2026 figures as a trend.

Review platforms: solid data for AI Overviews, thin for the rest

SE Ranking tracked 23 review platforms, mostly B2B software sites plus Yelp. At least one appeared in 34.5% of AI Overviews. Query wording mattered a lot: 49% for explicit review searches, 39.4% for “software” or “tools” searches, and 17.1% for “best” or “top” searches, where AI Overviews leaned on review blogs, media sites and ranking-style pages instead. Five platforms took 88% of review-platform links: Gartner Peer Insights (26.0%), G2 (23.1%), Capterra (17.8%), Software Advice (12.8%) and TrustRadius (8.3%). Review platforms were only 8.5% of all links, yet three of them ranked among the five most-cited domains.

Outside Google, the public numbers are older or use other metrics. In Profound’s 2024 to 2025 data, G2 held 1.1% of ChatGPT citations, and Yelp (0.8%), TripAdvisor (0.6%) and G2 (0.6%) appeared in Perplexity’s.

Grok’s source list leans heavily on user platforms. Ahrefs’ June 2026 list, built from more than 1.9 million US queries in Brand Radar, ranked Reddit first at 16.3% mention share, YouTube second at 15.1% and Facebook third at 13.9%, with Instagram 5.9%, Quora 5.5%, Amazon 5.0%, X 1.4%, Yelp 1.0%, TripAdvisor 0.9% and Trustpilot 0.8% further down. Mention share is a share of the top 50 domains’ combined citations, not of all citations, so these numbers run higher than Otterly’s and should not be set side by side.

We found no 2026 study that measures Trustpilot, Yelp, TripAdvisor or Amazon reviews across all engines on the same prompts. For consumer review sites, your own prompt data is the only reliable read.

Google’s UGC Fresh Data Program: what it says and what it does not

On October 8, 2026, Google documented the UGC Fresh Data Program, which it describes as “an initiative designed to help platforms display fresh, high quality, and authentic user-generated content.” It is open to platforms that “primarily host UGC, social postings, or forum discussions,” with public pages, active moderation and a user reporting mechanism. Approved platforms get a dedicated ingestion pipeline: content “ideally sent within minutes,” engagement counters updated “within 72 hours of creation.” Google promises a response to applications “within 6-8 weeks” and says the program “doesn’t guarantee that content will appear in Search.”

Two limits matter for brands. The page refers to “various features on Search” and does not name AI Overviews or AI Mode. It also names no participating platforms. Unless you run a community platform, this is not something you apply to. Our read: if large forums join, the likely effect is fresher threads reaching Google faster, which raises the cost of leaving a wrong or stale claim about your product unanswered.

Where to invest, engine by engine

Decision map for community and review investment by engine, with ethical guardrails

The rules below are Rankry’s reading of the studies above, not a separate study. Check them against your own prompts before moving budget.

  1. Start with the engines your buyers use. A team whose buyers live in Google’s AI surfaces and a team whose buyers ask Claude need different plans. Measure UGC share in your own prompt set first.

  2. For AI Overviews and AI Mode, show up where the threads and videos are. Answer existing Reddit threads that already rank for your buyer questions, publish YouTube videos that answer them, and keep review profiles complete for review-intent queries. The Reddit method is in our guide to getting cited on Reddit.

  3. For Perplexity and Grok, keep Reddit in the mix. Reddit still held 2.043% of Perplexity’s citations on September 20, and Grok’s June list was led by Reddit, YouTube and Facebook.

  4. For ChatGPT, Claude, Gemini and Copilot, put owned and editorial pages first. User content is a small share there. In Otterly’s data, comparisons were 16.1% of ChatGPT citations and listicles 78.2% of Claude’s, so invest in comparison pages, docs and placement in credible third-party roundups.

  5. Earn reviews within the FTC rule. Since October 21, 2024, businesses may not create, buy or sell fake reviews, offer incentives conditioned on positive or negative sentiment, post undisclosed insider reviews, present a site they control as independent, or suppress negative reviews through threats or false accusations. The FTC can seek civil penalties against knowing violators. Ask real customers, and make no incentive depend on what they say.

  6. Follow Reddit’s own rules. Rule 2 asks users to “participate authentically” and not to spam or manipulate content; Rule 5 bans intentionally misleading others or impersonating an individual or entity in a deceptive manner. Disclose your affiliation, use one real account, and never buy votes.

  7. Re-check every month. In Otterly’s data, ChatGPT’s Reddit share fell 94% from July to September 2026. Track citations per engine and act on shifts that last several weeks, as covered in how to track citations and sources in AI search.

Measuring your own mix

Every study here uses its own prompts, industries and definitions, and none of them is your category. Rankry runs a stable set of buyer-intent prompts every day across ChatGPT, Claude, Gemini, Perplexity, Grok and Google AI Mode and AI Overviews, and records which sources each engine cites when it mentions, ranks or recommends you. That shows directly whether Reddit threads, review profiles or your own pages carry your visibility on each engine, competitor comparison shows where rivals appear and you do not, and a new question is answered as soon as you add it.

FAQ

Which AI engine cites Reddit the most? Google’s AI surfaces, in OtterlyAI’s seven-engine data. OtterlyAI measured Reddit at 5.433% of AI Overviews citations and 5.264% of AI Mode citations on September 20, 2026, ahead of Perplexity at 2.043% and Gemini at 1.295%. Ahrefs’ June 2026 list also ranked Reddit first among Grok’s sources, using a different metric that is not directly comparable.

Does ChatGPT still cite Reddit? Rarely. In OtterlyAI’s data, Reddit held 3.765% of ChatGPT citations from July 1 to August 7, 2026, then 0.227% across September after a drop that began on August 8. In Otterly’s August study, official brand and competitor pages gained in 10 of 15 reports. The cause is unconfirmed, and OpenAI had not published an explanation as of that study, so check your own prompts monthly.

Do AI engines cite review sites like G2 or Trustpilot? Google AI Overviews does, mostly for review-intent queries. SE Ranking found at least one review platform in 49% of AI Overviews for explicit review searches, with Gartner Peer Insights, G2, Capterra, Software Advice and TrustRadius taking 88% of all review-platform links. For other engines, public data is older or uses different metrics, so measure your own category’s prompts.

Does Google’s UGC Fresh Data Program affect AI Overviews? Google has not said so. The program documentation, updated October 8, 2026, refers only to “various features on Search,” names no participating platforms, and says inclusion is not guaranteed. It gives approved forums and social platforms a dedicated ingestion pipeline, with content ideally sent within minutes. Whether that reaches AI Overviews or AI Mode is not stated, so treat any such claim as speculation.

Can I post reviews or Reddit comments about my own product? Only openly. The FTC rule in force since October 21, 2024 bans fake reviews, undisclosed insider reviews and incentives tied to positive or negative sentiment. Reddit’s rules require authentic participation and ban content manipulation and misleading others. Disclose that you work for the company, answer the question honestly, and let real customers write reviews.

Sources

  1. Why AI Engines Cite Different Sources but Recommend the Same Brands, BrightEdge, April 24, 2026
  2. What Content Do AI Engines Cite? A Study of 617,985 Citations Across 7 Engines and 5 Industries, OtterlyAI, October 14, 2026
  3. Review platforms lost 90% of traffic over the past 2 years, yet they top AI Overview citations for commercial queries, SE Ranking, January 29, 2026
  4. The September Update: The Month AI Search Stopped Agreeing About Social, OtterlyAI, September 22, 2026
  5. ChatGPT cut Reddit citations by at least 73% in August 2026. Here is what replaced it., OtterlyAI, August 27, 2026
  6. The Most-Cited Domains in AI: A 3-Month Study, Semrush, November 10, 2025
  7. AI Platform Citation Patterns: How ChatGPT, Google AI Overviews, and Perplexity Source Information, Profound, June 5, 2025 (updated August 2025)
  8. The 50 Most-Cited Websites in Grok (June 2026), Ahrefs, June 1, 2026
  9. UGC Fresh Data Program, Google Search Central, last updated October 8, 2026
  10. Federal Trade Commission Announces Final Rule Banning Fake Reviews and Testimonials, FTC, August 14, 2024
  11. Trade Regulation Rule on the Use of Consumer Reviews and Testimonials, Federal Register, August 22, 2024
  12. Reddit Rules, Reddit Inc., current policy page

See which forums, review sites and community threads each AI engine cites for your buyer prompts, every day. Start a free 7-day Rankry trial, no card required.

Enjoyed this article?
Share it with your network

Track your AI visibility

See how your brand appears across ChatGPT, Claude, Gemini, Perplexity, Grok, Microsoft Copilot, and Google AI Overviews.

Try Rankry