Agentic commerce is usually described as a transaction problem: how an AI agent will find a product, compare options, complete a payment, and move on. That framing misses the harder problem for brands. Before an agent can buy on a customer’s behalf, it has to understand what the brand stands for, what it sells, and why one option should be chosen over another.
That makes storytelling part of the commerce infrastructure. In an Adweek analysis, former Marriott and MLB CMO Karin Timpone argues that brands need to prepare for a world where AI agents mediate more of the customer relationship. The shift is not simply from human shopping to automated shopping. It is from persuading a person directly to being legible to the system acting for them.
AI agents will not just process product data
Timpone compares the moment with the early development of social media. At Marriott, she helped build social media hubs that made brand storytelling more visible, repeatable, and connected to the speed of culture. The lesson was organizational as much as creative: companies had to bring new skills and teams into the brand instead of treating social as a small channel extension.
Agentic commerce creates a similar pressure, but with a more consequential audience. Marketers are not only communicating with consumers. They are also influencing the agents that interpret consumer preferences, evaluate products, and potentially complete purchases.
That is why Timpone describes the opportunity as “Creative Data 2.0.” The phrase is slightly grand, but the underlying point is practical. Product information, customer insight, brand positioning, availability, pricing, and storytelling may need to work together in connected loops. If the information is incomplete or outdated, the agent may not simply show a less attractive result. It may fail to understand the brand at all.
The change also challenges the old separation between brand and performance teams. A campaign can create awareness, while a commerce system handles conversion. An agentic system may connect those stages, using an understanding of the brand to shape which products are recommended and which purchase path is offered.
That possibility is already raising a broader question about the future of advertising and discovery. AI agents could disrupt the web’s advertising economy because they may reduce the number of pages, searches, and product comparisons a person sees before making a decision. The brand still needs to be discoverable, but the route to discovery may no longer pass through a visible ad or a conventional search result.
The first decision is what role the brand wants to play
The most useful part of the Adweek framework is its attempt to give brands a starting point. Timpone identifies three possible roles for agentic commerce.
An advisor helps people navigate choices. A beauty brand could explain a reformulated product or guide someone toward an alternative. This requires timely information, especially when product details change.
An assistant anticipates a use case and assembles relevant options. The example is occasion-based retail, such as organizing food, décor, and beverages for a party. The brand is no longer presenting a catalogue alone. It is structuring its products around the customer’s situation.
An operator handles more of the journey. Timpone compares this with an airline itinerary that brings together dates, locations, prices, activities, and booking. In this role, the agent needs enough reliable information to do the operational work, not merely surface a promotion.
Those roles have different implications for content, data, measurement, and internal ownership. A brand that wants to be an advisor needs authority and clarity. An assistant needs organized product information and a useful understanding of context. An operator needs dependable systems for availability, payment, sign-on, and fulfillment.
The strategic consequence is straightforward: brands cannot treat agentic commerce as a feature to be added after the marketing plan is finished. They need to decide how they want to be understood before an agent decides whether they are relevant. The companies that win attention in this environment may not be the ones with the loudest campaigns. They may be the ones whose story, product truth, and transaction logic are clear enough for both people and machines to act on.