G2, Capterra and AI citations: what the data shows

Short answer

Review platforms are a real source for AI answers about software, mostly when buyers compare and evaluate. A complete, current profile is worth maintaining. Chasing review volume to win AI visibility is not supported by the data.

Review platforms matter more at the bottom of the funnel

In an analysis of about 35,000 ChatGPT citations (US only, December 2025) published on G2’s Learn Hub, review platforms accounted for 7.4% of citations on discovery prompts and 13.2% on evaluation prompts, “roughly 1.8× discovery’s 7.4%” ^[1]. Overall, review platforms were 8.6% of citations, against 70.4% for vendor and other sites.

Within review platforms, the same study found “G2 ecosystem (g2.com + acquired brands): 84%”, with gartner.com at 13% and trustradius.com at 1% ^[1]. G2 owns Capterra, Software Advice and GetApp, and published the analysis on its own site.

More reviews, slightly more citations

A second G2-published analysis covered 500 G2 categories and 30,000 citations: “Categories with 10% more reviews have 2% more citations”, and “reviews explain less than 2% of the variance” ^[2]. The author flags the cross-sectional design as a limit. Read plainly: review count is a small piece of the puzzle.

What G2 says about itself

G2 states that its LLM referral traffic “has increased by 10x year-over-year” and that it is “the #1 cited source for B2B software queries across all major LLMs” ^[3]. In March 2026, it announced an MCP integration that lets AI platforms “reference verified buyer reviews” directly from G2 ^[4]. These are company claims without a published method.

G2’s buyer survey also found that “85% of buyers think more highly of a vendor when AI includes them in an answer” ^[7], which says more about the value of being named than about review platforms specifically.

Independent studies found much lower shares

Two smaller 2026 studies by agencies looked at B2B software discovery questions:

• DerivateX: across 233 ChatGPT software recommendations, review aggregators accounted for 0.9% of citations, and G2 and Capterra each received zero ^[5].
• Growfusely: G2, Capterra, Gartner and every other review site combined were 7.3% of 1,739 citations, via the OpenAI and Perplexity APIs ^[6].

Both focus on discovery questions, which is consistent with the first finding: review sites weigh less at discovery and more at evaluation.

What to do

• Keep profiles complete and consistent with your site: category, integrations, company size served, compliance, pricing model.
• Ask for specific reviews. A review that names a use case and an integration answers an evaluation question; “great product” does not.
• Do not stop there. At discovery, comparison articles and vendor pages carry more weight. See the ChatGPT playbook.
• Measure by intent. Split your buying questions into discovery and evaluation, and check which sources the engines cite for each.