For over a year now, a new discipline has been showing up in digital agencies’ service offerings: GEO, short for “Generative Engine Optimization.” It’s positioned as the next evolution of traditional SEO, and the pitch is built on urgency: AI is moving at breakneck speed, so you need to act now.
The problem is that the closer you look at what’s being offered, the more you find the same recommendations and core discoverability principles we’ve been applying in SEO for years. In May 2026, Google published a guide dedicated to optimizing for its generative features, explicitly aiming to debunk some common misconceptions about AEO and GEO. Its position: SEO best practices remain the best way to improve your presence in generative experiences. In June, Google went even further, clarifying that llms.txt, content chunking, AI-specific rewriting, and special schema aren't needed for its generative AI features.
Is GEO really the next big thing?
Personally, I don’t think so! While the industry is busy figuring out how to better present our web pages to AI, a much deeper transformation is underway: AI won’t stay passive, content to simply read your website. It’s going to want to interact with it.
That’s where a new standard called MCP, the Model Context Protocol, comes in. MCP offers a standardized way to make a system’s data and capabilities accessible to AI agents. In other words, rather than letting AI navigate our interfaces blindly, we can give it the means to interact directly with the data behind them. You could think of MCP as an API designed from the ground up for AI agents. We’re no longer just talking about optimizing content so AI can understand it better. We’re starting to design platforms it can actually use.
So how do you optimize your website for AI today?
For discoverability, the answer is far less dramatic than some would have you believe: keep applying proven SEO best practices. Create relevant, original content. Structure your information properly. Make sure your pages are accessible, fast, and easy for search engines to understand. There’s probably no need to reinvent your digital strategy every time a new acronym comes along.
But if the real question is how to prepare your business for the arrival of AI agents, you need to ask what an agent could accomplish if it had access to your data and services. The answer will obviously differ from one business to the next, and that’s exactly what makes the topic so interesting.
GEO is mainly about optimizing how AI finds and understands what we publish. MCP opens up a much broader conversation about how AI will interact with your entire digital ecosystem. This time, it may no longer be about optimizing your website, but about rethinking its purpose and what AI will be able to do with it.