Letter to the Board #1 : Tomorrow, your first customer won't be human

For decades, businesses have learned how to win over consumers. We have refined marketing, design, advertising, storytelling, and customer experience. We have studied cognitive biases, built iconic brands, partnered with influencers, and optimized our visibility on Google. An entire industry has been built around one fundamental question: how do humans make decisions? That era is coming to an end.

Within a few years, a growing share of our decisions will no longer be made directly by people, but by their AI agents. Welcome to the age of algorithmic capitalism.

Booking a hotel. Choosing an insurance policy. Comparing energy providers. Buying a laptop. Planning a business trip. Selecting a supplier. We will no longer ask, "What are my best options?" Instead, we will simply tell our AI agent, "Take care of it." Before long, we may not even review the final decision.

The end of marketing...as we know it

AI agents will analyze thousands of variables, compare competing offers, negotiate prices, complete transactions, and monitor delivery or execution. At first glance, this may seem like another technological evolution. It represents a fundamental shift in the way markets operate. The primary customer of a business will no longer be just a human buyer. It will increasingly be the artificial intelligence acting on that person's behalf. That distinction changes everything.

Today, brands win because of their image, their story, and their ability to create emotional connections. Tomorrow, those factors will still matter in some industries, but they will no longer be sufficient when an AI agent is making the purchasing decision. AI has no nostalgia. It does not aspire to belong. It is not persuaded by an award-winning advertising campaign or by elegant packaging. Instead, it evaluates measurable signals: product quality, reliability, delivery times, cybersecurity, total cost of ownership, regulatory compliance, environmental impact, reputation, and the ability to deliver on commitments. In other words, we are moving from an economy of persuasion to an economy of proof.

Companies will have to learn how to speak to AI

For the past twenty years, businesses have learned how to be discovered by search engines.

Tomorrow, they will need to be understood, evaluated, and recommended by AI agents. Search optimization will evolve. Corporate websites will evolve. Content strategies are already changing. More importantly, companies themselves will have to become machine-readable.

They will need to expose reliable, structured, and verifiable information about their performance, guarantees, cybersecurity policies, environmental commitments, regulatory compliance, traceability, and service quality. In other words, companies will need to convince two customers at once: the human... and the human's AI agent.

Sales teams will also face a new kind of counterpart: the customer's AI. This raises a question that should already be on every board's agenda: what must we do today to become the natural recommendation of tomorrow's AI agents?

A much more political battle than it appears

The answer has two dimensions. The first concerns companies themselves. Businesses will need to rethink their strategy, communication, and digital presence to ensure they are identified as the preferred option by AI agents. The second is even more strategic: who will decide the criteria these agents use?

Imagine your AI assistant choosing your next car. Should it prioritize price? European manufacturing? Carbon emissions? Cybersecurity? Human rights? Repairability? Supply chain transparency? Technological sovereignty? None of these choices is neutral. Whoever defines the criteria ultimately shapes the market.

Will those priorities be determined by OpenAI, Anthropic, Gemini, or another model developer? Will users be able to customize them? Or will governments eventually regulate them? Behind every AI recommendation already lies a complex set of economic, ethical, and geopolitical choices.

Large language models are no longer simply productivity tools. They are becoming economic intermediaries capable of directing billions of euros in consumption and investment. The next competitive battlefield will therefore not be artificial intelligence itself. It will be algorithmic trust. Tomorrow's winners will not simply be the companies that know how to persuade consumers. They will be the companies that artificial intelligence considers the most trustworthy.