Trust First Agent Monetization
Instinct
The key constraint on agent monetization is that the assistant has to behave like a buyer’s agent, not a lead generator. Once an assistant can compare flights, reserve restaurants, order groceries, and complete checkout, even small ranking bias changes what gets bought. That makes trust more important than short term affiliate yield, because the product only becomes a habit if users believe it is optimizing for price, fit, convenience, and reliability on their behalf.
-
Travel and commerce businesses already earn affiliate and booking fees, but they do it inside products where users expect marketplace incentives. Super.com mixes membership, card interchange, travel commissions, and retail affiliate revenue. Ramp adds affiliate fees around business travel. That model works when the product is clearly a seller or intermediary, not a personal representative.
-
Checkout cross sell businesses like Rokt show the opposite logic. They make more money by inserting offers after a transaction, like hotels after a flight booking. That can lift monetization, but it trains the system to maximize attach and take rate, not necessarily the best outcome for the user’s original task.
-
The strongest external signal is that leading AI shopping products are emphasizing independence from merchant influence. OpenAI says shopping results in ChatGPT search are not ads and are not influenced by partnerships. That framing reflects how fragile trust is when an assistant starts choosing products and merchants inside the buying flow.
This pushes the category toward subscription first economics, with referral revenue used carefully in places where user intent is already explicit and merchant selection is low stakes. The winners in agentic commerce will be the products that can monetize around the transaction while keeping recommendation logic visibly aligned with the user, because that trust is what unlocks repeat delegation.