Agent Commerce per MIT Tech Review: why AI agents require data, not banners

AI agents are transitioning from returning lists of links to actually executing tasks: booking reservations, making purchases, comparing offers. But for this to work reliably, they need two key resources: accurate structured data and complete user context.

From Prompt to Action

Imagine you tell a digital agent: "book a family trip to Italy, use my bonus points, stay within budget, and pick hotels we've liked before." Instead of returning a list of links, the agent independently builds the itinerary and completes the purchase. This transition—from assistance to execution—is the essence of agentic commerce.

MIT Technology Review observes that we are at the threshold of the next wave of e-commerce, where the primary interface is no longer a website or app, but an agent. The change appears evolutionary, but its implications for business are radical. Currently, OpenAI Operator, Google Shopping, and Claude's agentic modes are accelerating this transition. The technology is already ready—the question is whether retailers and services are prepared for it.

Truth Matters More Than Beauty

Traditional e-commerce is optimized for humans: attractive product cards, emotional banners, visual personalization. None of this is relevant to an agent. An agent needs structured data: accurate prices, real inventory status, machine-readable shipping and return conditions. It doesn't need a "brand story" on the homepage—it needs a reliable API or a clean data feed. If the data is outdated or contains errors, the agent will either make an incorrect decision or refuse to work with that source altogether.

• Real-time price and inventory accuracy is mandatory
• Structured product attributes matter more than marketing descriptions
• Machine-readable shipping and return conditions must be directly accessible
• Inaccurate data leads to incorrect transactions or complete agent refusal

This shifts investment priorities: money flows not into UX design, but into data quality and accessibility. Companies that continue investing only in visuals risk becoming invisible to agents.

Context Determines Choice

The second component is user context. An agent must know not only what to buy, but for whom, under what constraints, and considering which preferences. An effective agentic experience is built on accumulated history: which hotels were liked before, what budget is comfortable, which airlines to avoid. This requires user trust in the system—and a serious approach to personal data management.

An agent without context will revert to the same behavior as a regular search engine.

"Agentic commerce runs on truth and context"—this is how MIT

Technology Review articulates the fundamental condition of the new era of shopping.

For companies, this means a new challenge: they must not only create a good product but also actively structure their offering so that agents can interpret it correctly and explain to the user why this particular option is the best fit.

New Rules for Business

The transition to agentic commerce breaks the familiar sales funnel. In the past, brands spoke directly to consumers through advertising, content, and UX. Now an agent stands between them—and it is the agent that decides which vendor makes it into the user's final choice.

• SEO yields to agent-first optimization—structured data and open APIs
• Brands compete not for clicks, but for agent trust and data-feed quality
• Marketing budgets shift toward data infrastructure
• Companies that first adapt their data for agent consumption will gain disproportionate advantage

What This Means

Agentic commerce is not distant future: the tools already exist, major platforms are actively implementing them, and the first cases with real transactions are already being published. Brands that learn to speak the language of agents—structurally, precisely, in real time—will win the next era of e-commerce. Those who continue to optimize only for the human eye risk being left behind.