
A Spanish family-owned florist, Persa Flores, has discovered that artificial intelligence is now a substantial contributor to its revenue stream. After analyzing orders generated through ChatGPT, the company found that approximately one in five online purchases originated from this chatbot.
Uncovering the Trends
José Pérez Sánchez of Persa Flores explained in an email, “Your 22 October column, ‘AI Shopping Metrics Considered‘, and ‘Mixed Reports on AI Ecommerce Traffic‘ from April are why I’m writing,” which prompted him to share his company’s experience. He revealed that since September 2026, one in five of their online orders has been attributed to customers who initially found them through ChatGPT.
Pérez Sánchez utilized his company’s data to build upon the arguments presented in the earlier articles, highlighting the challenges of identifying AI-referred shoppers without order-level attribution. To address this, he outlined how Shopify records detailed information for each order, including the customer’s first and last visit before making a purchase, the referring site, the landing page URL, and the date.
According to Pérez Sánchez, an order is considered ChatGPT-driven if either the first or last visit came from ChatGPT or OpenAI, and every such order arrives via a link tagged with utm_source=chatgpt.com, which ChatGPT automatically adds. However, he noted that only about one in eight of these orders also list ChatGPT as the referrer, meaning that relying solely on referrer data would cause most of this traffic to be overlooked.
The term “referrer” can be confusing, partly due to its traditional misspelling as “referer.” In technical terms, a referrer refers to the last URL a visitor came from before arriving at a target website, which is passed by the web browser in the HTTP header and read by analytics tools. However, these tools can fail to capture the referrer due to various reasons such as policies, privacy settings, or protocol downgrades, leading to the traffic being mislabeled as “direct” or “unknown,” as was the case with Persa Flores’ ChatGPT visits.
Organic Growth
After identifying the significance of ChatGPT-driven orders, the marketing team at Persa Flores observed a notable increase in the channel’s importance. ChatGPT accounted for 4% of the florist’s online orders in July 2026, rising to 12% in August, 21% in September, and reaching 23% by October 27. Similarly, ChatGPT contributed 2% of online revenue in July, 11% in August, 18% in September, and 26% by October 27.
Role of Existing Content
The city pages explain who Persa Flores is and show available bouquets with delivery prices. Each page focuses on sending flowers to a specific city in Spain. These pages, available in Spanish and English, were already in place before the growth in ChatGPT traffic.
Search optimizers have long used location- or intent-specific landing pages to match a search query more closely than a category or home page. Pérez Sánchez wrote that when someone asks ChatGPT how to send flowers to a town, a page dedicated to exactly that is an easy answer to give. ChatGPT appears to match a shopper’s question to a page with the answer.
New Customer Acquisition
Having identified ChatGPT as a significant source of orders, Pérez Sánchez then examined the behavior of these shoppers. Every order attributed to ChatGPT came from a first-time customer, as indicated by Shopify’s records showing no previous orders from the customer’s email address or account. Most of these customers made a purchase within a day of their first recorded visit, with about half of them paying from outside Spain, typically to send flowers to someone in the country.
The average order value from ChatGPT customers was similar to that of other online orders, although Pérez Sánchez noted that the company does not track conversion rates by source with sufficient precision to compare ChatGPT directly with Google or other channels. The order data revealed that ChatGPT was not only bringing in new customers but also introducing new paths to purchase. As Pérez Sánchez wrote, “84% of ChatGPT customers first arrived on a city page, compared with about a third of customers from Google.”
