Search patent of the week: ChatGPT Referrals to E-Commerce Websites

This episode is focussing a research provides an empirical evaluation of Organic Large Language Model (OLLM) traffic, specifically analyzing referrals from ChatGPT to nearly one thousand e-commerce websites over a year. Contrary to high industry expectations, the study finds that OLLM traffic generally underperforms traditional digital channels, including organic and paid search, in critical financial metrics. For instance, the channel delivers significantly lower results for both Conversion Rate (CR) and Revenue Per Session (RPS) than most competitors. While the channel’s high relevance is indicated by a favorable Bounce Rate, any gains in conversion over the year were negated by a declining Average Order Value (AOV), suggesting OLLM drives smaller, lower-priced purchases. Consequently, marketers are advised to treat OLLM as a long-term discovery tool and optimize Product Detail Pages (PDPs) as standalone landing pages using structured, granular product data to aid LLM synthesis.https://www.kopp-online-marketing.com/patents-papers/chatgpt-referrals-to-e-commerce-websites