{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"The GEO Show","title":"AI Is Talking Buyers Out of Purchases","html":"<iframe width=\"100%\" height=\"180\" frameborder=\"no\" scrolling=\"no\" seamless src=\"https://share.transistor.fm/e/050f8219\"></iframe>","width":"100%","height":180,"duration":1080,"description":"Welcome to episode 23 of The GEO Show, the briefing on generative engine optimization, LLM citations, and AI search visibility. Brought to you by GEOforge.\nIn this episode, Paris Childress, founder of Hop AI and co-founder of GEOforge, records a September 8, 2026 solo news roundup. The lead story is Semrush and Exploding Topics finding that 57.5% of AI users said chatbot information caused them not to buy, then nine more beats on mentioned versus chosen, single-check misclassification, Reddit citation collapse, five-engine source overlap, Google AI Mode citation architecture, production UI collection APIs, Claude retrieval versus citation, historical AI data moats, and Mastercard's agent shopping forecast.\nIn this episode:\n🛒 Semrush/Exploding Topics: AI talks buyers out of purchases\nA vendor-sponsored Semrush and Exploding Topics survey of 2,338 US adults (Sept 7) found 57.5% of AI users said chatbot information caused them not to buy. 65% said AI had at least partially replaced Google for product research. Among weekly AI users, 73.6% bought from an organic AI recommendation. Paris's GEO take: purchase exclusion, reputation, and buyer risk now sit beside discovery.\n📌  Latent Space: mentioned is not chosen\nLatent Space's Frontier AI visibility tracker (6,762 answers, 161 categories) shows Kysely in 42/42 answers but first choice only once. GPT-5.6 Sol and GPT-6 Astra category leaders differed in 33/121 comparable categories. Mention share of voice can overstate commercial visibility; recommendation role and sentiment matter.\n🎲 Surfaced by: one AI check can misclassify visibility\nAcross 536 combinations with at least five checks, 80% of appearance cases were inconsistent; a single check matched the majority only 72.2% of the time on inconsistent sets. Paris: target roughly 30 to 40 runs per prompt before volatility flattens.\n📉 Surfaced by: ChatGPT Reddit citations 37.9% to 3.4%\nIn fixed cohorts across two 14-day windows, ChatGPT Reddit citations collapsed while...","thumbnail_url":"https://img.transistorcdn.com/7cDHOTKwTPJbamMzkSU5XMeYqbfRSPk5nMQ6zhbTeaI/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS9kY2Ri/MGIyZjViNzJiODc5/ZGM5MmJhYmE4NmYx/MTVhYy5wbmc.webp","thumbnail_width":300,"thumbnail_height":300}