AI agents learn about our preferences, budgets and values. They analyse millions of data points, prices, inventory status, customer reviews, sustainability and lead times, and make decisions on our behalf. Unlike Google, which still favours merchandisers by indexing the store offering, AI agents are on the side of consumers. They don’t care about marketing budgets, but are optimising in the best interest of the customer.
An AI agent isn’t just a chatbot. It acts independently, compares alternatives, negotiates prices and makes purchases. For consumers, this means convenience. For retailers, it poses an existential challenge: how do you build loyalty when customers no longer shop themselves?
When purchases are made machine-to-machine, the emotional narratives lose their power. Brands that today compete on image and experience must instead win the battle to be machine-readable. Agents do not act on emotion; they choose the option that the data supports.
The decisive factor will be how well your offering is structured and understandable to an algorithm. Those who cannot demonstrate their value chain with transparency risk being left behind. Merchants who fail to adapt risk being reduced to background tools in agent-driven marketplaces.
Don’t be passive!
The biggest threat isn’t AI agents, but passivity. Agents will soon become an integral part of the shopping experience, streamlining decisions and pushing down prices. The question isn’t whether to work with agents, but how. The retailers need to choose a path. Of the big giants, Walmart, for example, has started collaborating with OpenAI, while Amazon is choosing to develop its own system of AI agents entirely.
Regardless of strategy, it is necessary to understand which customer problems the agents are supposed to solve and to be able to build trust.
An agent that gives the wrong answer quickly loses credibility, but companies that shape how agents meet customers early on will be the strongest competitors.
Three steps to winning the AI agent race
- Choose your AI agent strategy: AI agents are changing the entire purchasing process. Brands must choose a strategy: to collaborate with third-party agents such as OpenAI or Google or build their own trained AI agent using their own data. Both strategies work, but the choice determines whether you retain the customer relationship or hand it over to the agent.
- Understand your customer’s problem: What needs should the AI agent solve? Simplified everyday shopping, personal advice or price optimisation? When agents make decisions for people, trust, transparency and integrity become the new currencies of loyalty. The customer must be able to trust that both your agent and your data reflect the customer’s values.
- Build your data foundation and experiment: It all starts with organising your data. Product information, prices and delivery times must be understandable to algorithms. But it’s also about organisation and culture, creating ways of working where agents and people collaborate. Those who dare to test and learn on an ongoing basis will be in the lead.
The first steps are already being taken. Google is testing automatic purchase flows via Gemini. Amazon is letting agents to buy from other websites, and OpenAI is already handling the entire chain from comparison to payment.
The question isn’t whether commerce will change, but who will be able to build the foundation in time. Companies that are already making their brands machine-readable, ensure that their offerings are understood by agents, and get to know their customers’ digital twins will not only survive. They will be leaders in what is perhaps the most transformational change in commerce since the internet.