The Economics of Agentic Commerce

New commerce channels often appear most attractive before their full economics become visible. Revenue arrives first. Fees, advertising, returns, platform dependence and operating complexity become clearer later. Brands have seen this pattern with marketplaces, paid social and retail media, and agentic commerce deserves the same financial discipline from the beginning.

The question is not whether AI shopping agents generate revenue. The question is what kind of revenue they generate and how much economic value remains with the brand.

Customer acquisition cost may move rather than disappear

AI agents could reduce some acquisition costs by improving product matching and reducing paid persuasion. It would be risky, however, to assume valuable customer attention will remain unmonetized. Successful intermediaries have strong incentives to capture part of the value they create. Brands may eventually pay through transaction commissions, referral fees, promoted recommendations, preferred access, platform charges or integration costs. The terminology may change; competition for access to high-intent customers will not.

Attribution will become more difficult

A shopper may learn about a category on social media, ask one AI service for recommendations, use another service to compare prices and complete the transaction through a retailer selected by the agent. Assigning the entire sale to whichever system records the final transaction would overstate its contribution. Agentic commerce adds more intermediaries to a problem that already exists in multichannel ecommerce.

Management should therefore focus on incrementality and channel economics rather than attempting to assign perfect credit to every interaction.

Contribution matters more than agent-generated revenue

A useful board-level framework begins with net revenue but quickly moves to gross margin, referral costs, promotions, advertising, fulfilment, returns and customer value. Two channels generating identical first-order revenue can create very different enterprise value if one produces a known repeat customer and another produces an anonymous transaction that must be reacquired.

The EVA framework on Ecommerce Metrics for Boards applies directly. Channel revenue should be evaluated alongside contribution, concentration, paid-demand dependency, inventory and returns.

Efficient comparison can increase price pressure

AI agents can reduce the cost of comparing products and offers. Brands with genuine differentiation may benefit because agents can identify attributes consumers would struggle to research manually. Brands whose economics depend partly on comparison friction may face more pressure. This creates another link between product information and economics.

New intermediaries can create new concentration

If one or two AI platforms account for a substantial share of discovery, brands can recreate familiar dependency. Management should monitor referral share, customer-data dependence, technical integration dependence and changes in bargaining power, not just revenue. This is conceptually similar to Amazon Concentration Risk.

The broader EVA principle also applies: growth should be evaluated on economic quality rather than headline revenue. See When Ecommerce Growth Creates Enterprise Value.

What changed in agentic commerce, and what it means for brands.

A concise weekly brief for brand executives navigating the shift from human-led shopping to agent-mediated commerce. Follow the developments that matter, the implications for brand control and governance, and the questions senior teams should be asking now.

Read The Agentic Commerce Brief on Substack