New analysis suggests that AI shopping agents like Meta's Muse may not generate enough transactions to be profitable. While proponents envision agents disrupting e-commerce by finding the best deals, a report highlights that the average consumer has limited opportunities for agent-assisted purchases, potentially limiting revenue streams for these services.
The potential profitability of AI shopping agents is uncertain if consumers have limited opportunities to use them for transactions, which could impact the e-commerce landscape and the business models of companies like Meta and Instinct.
Meta's vision for its AI agent, Muse, centers on consumers delegating their online shopping to the service, with Meta aiming to profit from transaction fees. This concept has fueled significant investment in AI agent startups, with Instinct recently achieving a $10 billion valuation.
However, analysts at MoffettNathanson present a more cautious outlook, arguing that the actual utility and adoption of such agents may be limited. Their analysis suggests that out of an average consumer's 47 monthly payments, only about 11 are realistically candidates for AI agent handling. This excludes essential automated payments like bills, fuel purchases, and in-person retail transactions.
The limited number of potential agent-assisted transactions raises questions about the profitability of services like Muse or Instinct, especially if they rely on small per-transaction fees. The report implies that for these agents to become truly disruptive forces in e-commerce, widespread adoption would be necessary, which may be challenging given the constrained opportunities for their application. One theory suggests that Muse might evolve into an advertising platform for Meta, while another posits a future where users pay subscription fees for highly useful agents.
Pick the topics you care about. Get only what matters, on your cadence.