Persa Flores, a family-owned florist in Spain, discovered that ChatGPT has become a significant source of online orders. Since September 2026, one in five of the company's online orders came from customers who found the business through ChatGPT. The share grew from 4% of orders in July 2026 to 23% by October 27, while ChatGPT-driven revenue climbed from 2% in July to 26% through late October.
The key insight is that ChatGPT traffic is difficult to identify using standard analytics. Although ChatGPT tagged visits with utm_source=chatgpt.com, only about one in eight of these orders showed ChatGPT as the referrer in Shopify's data. Most traffic was labeled as direct or unknown, causing merchants who rely on referrer data alone to miss the channel entirely. Every ChatGPT-attributed order came from a first-time customer, and 84% of ChatGPT customers first arrived on a city-specific landing page—compared with about a third of Google customers. Persa Flores' city pages, which focus on sending flowers to specific Spanish cities, proved to be an effective match for ChatGPT's recommendations, suggesting that familiar SEO practices like location-targeted pages may boost AI-driven discovery.
For commerce practitioners, this case demonstrates that AI-driven sales are growing but remain largely invisible in conventional traffic reports. Identifying ChatGPT as a source required order-level analysis using UTM parameters and customer visit history, not just referrer data. The florist's experience suggests that merchants should implement attribution methods beyond standard analytics tools to capture and optimize for AI-powered shopping channels.