![\"DerivateX](\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/06\/derivatex_logo.jpeg\"){kind=logo}
\n Vendors grading themselves are 51% of ChatGPT\u2019s software citations, trusted analysts just 16%. DerivateX calls it the Authority Inversion.<\/p>\n
\t\t\t\t
\n Vendors grading themselves are 51% of ChatGPT\u2019s software citations, trusted analysts just 16%. DerivateX calls it the Authority Inversion.<\/p>\n
When ChatGPT recommends business software, it leans far more on vendors\u2019 own websites and on blogs almost no one has heard of than on Gartner, G2, or the business press. That is the finding of a new study from DerivateX<\/a>, a B2B SaaS search agency, which traced every source ChatGPT cited across 40 software categories.<\/p>\n ChatGPT cited a small consulting firm\u2019s blog more often than Gartner. The institutions that vetted software for buyers have been replaced by vendors and sites no one has heard of.\u201d<\/p>\n \u2014 Apoorv Sharma, Co-Founder, DerivateX<\/p>\n Software vendors writing about themselves accounted for 51 percent of the sources ChatGPT cited. Small, often anonymous websites made up another 23 percent. The analyst firms, review platforms, and business press that have guided software buying for two decades accounted for just 16 percent combined. Also Read:\u00a0AiThority Interview with Matej Bukovinski, Chief Technology Officer at Nutrient<\/a><\/p>\n In several categories, sites with no public profile outranked household names. One product-management tool\u2019s blog, ideaplan, was cited six times, more often than Forbes, Reuters, or Gartner. In the marketing automation category, a single blog post from a small consulting firm supplied four of the recommended tools. DerivateX describes the pattern as the Authority Inversion: the trusted middle of software research has been hollowed out, and most buyers cannot see what replaced it.<\/p>\n \u201cFor twenty years we taught software companies that trust was earned slowly, from analysts, review platforms, and the press,\u201d said Apoorv Sharma, co-founder of DerivateX. \u201cAI quietly threw that hierarchy out. Most companies are still spending against a map that no longer decides anything.\u201d<\/p>\n The findings come from a new DerivateX study of\u00a0where ChatGPT gets its software recommendations. For each category, the company submitted one buyer-intent query to ChatGPT with web search enabled, then classified all 188 cited sources by type across 105 domains. DerivateX presents the work as a snapshot of one platform and one query type, the moment a buyer is closest to a purchase decision.<\/p>\n The study carries practical implications for both sides of the market. For buyers, it suggests treating AI software recommendations as a starting point to verify rather than an independent verdict, since the most-cited sources are frequently self-interested or unvetted. For software companies, it argues that the channels many teams court most, analyst coverage and review-site badges, are not the ones AI engines cite, and that visibility now depends on surfaces most companies do not track.<\/p>\n DerivateX helps B2B SaaS brands get found and cited inside ChatGPT, Perplexity, Gemini, and Claude. The company provides\u00a0LLM SEO\u00a0services and publishes buyer guidance, including a guide to hiring an\u00a0AEO agency for B2B SaaS\u00a0for teams evaluating outside help.<\/p>\n Also Read:\u00a0\u200b\u200b<\/a>AI systems \u2013 Interoperable AI systems: Connecting models across platforms<\/a><\/p>\n
The platforms most associated with software research barely registered. G2 and Capterra, the two largest software review sites, received zero citations across all 40 categories. Gartner appeared only twice, and both times through its user-review pages rather than its analyst research.<\/p>\n