“La Ecoserie,” a branded content success story for ECOCESTA

_{Estimated reading time: 5 minutes}

For more than two decades, digital marketing has been built on a relatively stable premise: that people search, compare, click, browse, leave data, and convert. For their part, brands design customer journeys, optimize landing pages, buy traffic, measure sessions, attribute conversions, and strive to improve every step of the funnel.

The agent-based web is coming, and it will change everything.

The agent-based web is the re-evolution of the internet. Delegating tasks to artificial intelligence agents capable of understanding intent, searching for information, comparing alternatives, making assisted decisions, and taking action on your behalf—in other words, it’s the dream of usability. Simply by stating your goal, the agent will handle part or all of the journey. For example:

“Find me an affordable hotel in Madrid for two nights, with flexible cancellation, good Metro access, and for less than 180 euros.”
“Buy this product if the price drops and it meets these conditions.”
“Contact 5 real estate agencies that have apartments with these features.”

In this scenario, the browser, search engine, comparison site, assistant, and checkout begin to blend together, and this has profound implications for all of us in marketing.

Until now, much of digital marketing has focused on attracting users to a company’s own assets. In the agency-driven web, part of the user journey can take place before, after, or even outside the scope of those assets.

Brands are no longer competing solely for clicks; they’re competing to be among the brands selected and implemented by an agent. And the sooner your brand is ready, the more business opportunities you’ll generate.

Why It Matters Now

Interest in agent-based AI is no longer just a theoretical concept; we must face it head-on and incorporate it into our list of priorities. Major consulting firms are beginning to quantify its economic and organizational impact with significant figures: McKinsey estimates up to $3–5 trillion in agent-mediated global commerce by 2030; Capgemini estimates an impact of 450,000 million by 2028; and Gartner predicts that one-third of interactions with GenAI will use agents to complete tasks.

The opportunity is enormous, but so are the doubts about how to embrace it. The consensus points to a gradual loss of control over the funnel and the data: Forrester warns that the customer relationship may shift toward the agent and that some of the information will remain at that level. Capgemini, Bain, and PwC agree that without first-party data, martech integration, redesigned processes, and connected systems, AI will not have a real impact.

The common message is clear: agentic AI requires not only the adoption of new tools, but also a rethinking of entire processes.

The moves by the big tech companies show that the agent-driven web is taking shape at a rapid pace.

Google continues to make progress on its Universal Commerce Protocol (UCP), with the goal of enabling purchases and reservations directly from conversational experiences such as AI Mode or Gemini, bringing user intent closer to conversion. It starts with a clear advantage: years of development in Google Shopping and Hotel Ads, along with feeds and integrations with retailers, hotels, and technology partners.

Other major tech companies are moving in the same direction. OpenAI is doing so with Operator, which can navigate, click, and perform actions on the web; Anthropic, with Claude’s “computer use” capabilities; and Microsoft, by integrating agents into Copilot, Office, Teams, Windows, and enterprise environments, as well as driving the “open agentic web.” Amazon is exploring this area with Buy for Me, which allows users to purchase products from other brands within its platform, while Salesforce, with Agentforce, is bringing agents into CRM, sales, customer service, and sales automation.

For marketing, the agency website can be just as disruptive and transformative as the advent of the internet, the smartphone, or social media were in their time.

Its main effect will be a loss of control over the customer journey: part of the decision-making process may take place without the user visiting the brand’s website, or only at very late stages of the journey.

This requires us to rethink the role of our own assets. The website will remain important, but it will no longer always be the centerpiece of the customer journey. Brands will have to optimize how agents understand, interpret, and recommend their value proposition.

There will also be a loss of data and visibility. If the user delegates more steps to an agent, brands may lose critical signals such as traffic, web behavior, drop-offs in the funnel, or sales attribution. Measurement will need to evolve toward more hybrid models, based on server-side events, persistent identifiers, conversion imports, and traceability beyond the click.

Integration will be key. The agent-driven web will reward brands that are easy for machines to understand and execute: those with structured information, APIs, feeds, product data, availability, prices, policies, inventory, brand identity, and clear business rules. Integration will become a marketing lever.

It is in this context that Agent Experience Optimization (AEO) emerges. Brands must ensure that their products, services, or loyalty benefits are clear to agents.

5 Areas to Focus On That You Can Start Working On Right Away

The agent-based web is still under construction, but waiting for it to be fully developed might mean arriving too late.

1. Audit your customer journey: Identify which steps could be handled by agents

2. Prepare the data: Product information, rates, availability, policies, content, benefits, and unique selling points must be structured, up-to-date, and interconnected.

3. Review the metrics: We’ll need to define new channels, new sources, new events, and new attribution models for customer journeys that don’t always go through the website.
   
4. Strengthen your first-party data: In an increasingly intermediary-driven web, recognizing customers and establishing a direct relationship with them will give you a competitive advantage.

5. Establish governance. Agents can perform tasks, but they can also make mistakes, misinterpret information, amplify biases, or create security and privacy risks. Autonomy requires limits, traceability, and control.