Angelfish study: AI cites third-party sources more than brands for B2B software answers

Angelfish Marketing analyzed 1,052,053 AI citations across six B2B SaaS brands and found that AI answers about software are built mostly from third-party sources, not company websites. The study suggests brands need visibility across reviews, communities and trade media to show up in AI recommendations.

Why it matters: - AI visibility for B2B software is not just a website problem. - The study suggests brands need presence across the sources AI assistants actually read, or they may miss out on being recommended to buyers. - The finding is especially important for SaaS companies competing in crowded categories where AI compares multiple vendors by default.

What happened: - Angelfish Marketing published original research on July 8, 2026, analyzing 1,052,053 AI citations. - The study looked at how AI assistants assemble answers about B2B software for six established SaaS platforms. - The brands covered CRM, project management and marketing automation, with two brands in each category. - The data came from the SEMrush AI Toolkit and blended results from ChatGPT, Google AI Overviews, Google AI Mode and Gemini.

The details: - Across answers that mention a brand, the brand’s own website appeared as a cited source only about 13% to 32% of the time. - Most of the answer material came from competitors, community platforms, review sites and trade press. - YouTube and Reddit were the two most-cited individual sources across the study. - LinkedIn ranked third among sources, ahead of every review site and publication. - Medium, review platforms such as G2 and Capterra, and trade titles such as TechRadar and Forbes appeared repeatedly across the brands studied. - In total, 42 third-party sources appeared consistently across all six brands. - The study found AI answers about B2B software are often built by comparison, not isolation. - When buyers ask questions like the best CRM for a small sales team, AI weighs multiple products against one another. - That pattern pulls in sources that discuss entire categories, not just single brands.

Between the lines: - The research draws a sharp line between being eligible to appear in AI answers and being well represented in the answer itself. - Owned content and technical SEO help a brand get cited at all. - Third-party coverage, reviews and community discussion appear to do the heavier lifting once AI assembles the response. - Angelfish’s view is that many growing brands have focused on the first job and neglected the second. - This is descriptive research, not causal research, so it maps citation patterns rather than proving what drives them.

What’s next: - Angelfish says the next step for brands is to build visibility across the third-party sources AI systems trust most. - The company points to reviews, communities and earned media as the places where answer engine optimisation can influence visibility. - The full study, including the ranked list of the sources AI cites most for B2B SaaS, is available in Angelfish Marketing’s report. - Angelfish Marketing also notes its work in SEO, AEO and paid media for B2B companies in SaaS, technology and professional services.