OpenAI's Shopping Pivot Is a 'Disaster'

The push for "conversational commerce" hinges on the idea that AI can remove friction from online shopping. However, early attempts, like Amazon's voice shopping on Alexa, demonstrated that consumers don't always want a purely efficient experience; only 2% of Alexa users made a voice purchase in its early years, and 90% of those never made a second. The failure of AI shopping assistants often boils down to misunderstanding user intent and the psychology of shopping. Roughly 60% of online shopping involves "hedonic behavior" or browsing for discovery and entertainment. AI, by design, often eliminates the "friction of wondering," which is a key part of the dopamine loop that builds anticipation and desire for a product. From a marketing analytics perspective, the failure would be visible in key performance indicators (KPIs). A clunky AI shopping experience would likely lead to a high shopping cart abandonment rate, which averages nearly 60% for U.S. online shoppers, and a low conversion rate, the percentage of visitors who make a purchase. These metrics are fundamental for building portfolio projects in SQL or Tableau to analyze e-commerce performance. Independent analysis highlights the structural challenges AI agents face in gaining the contextual knowledge that influences purchases. An AI might not understand a family's specific dietary preferences, leading to rejected meals, or fail to recognize that a pantry item has already been purchased, resulting in duplicates—costly errors that erode user trust. The problem extends beyond user experience to the business model itself. Retailers are motivated to control their customer relationships and monetize their first-party data through advertising. An intermediary AI shopping agent disrupts this direct relationship, making widespread merchant integration a significant hurdle. This pivot is symptomatic of a broader "hard floor of reality" hitting the AI industry. OpenAI's monthly revenue reached $1 billion by July 2025, but the company is burning through $8 to $12 billion annually. With only about 5% of its 190 million daily users paying for subscriptions, the company is forced to explore new revenue streams, even if they are not a natural fit for the core product. Successful AI in retail often focuses on augmenting, not replacing, the shopping journey. Sephora's "Virtual Artist" app uses AI and AR to allow virtual makeup try-ons, directly addressing a key barrier to online cosmetic sales and increasing conversion rates. This serves as a contrasting case study on leveraging AI to enhance, rather than eliminate, the customer's decision-making process.