How is generative AI (ChatGPT, Copilot or Gemini) impacting traditional SEO, and what does GEO mean for brands?
The shift is quite fundamental: we are no longer competing simply to appear on Google, but to be the answer that the AI decides to provide.
Previously, the model was very clear: the consumer searched online, Google displayed results, and brands competed for the click. Now the behaviour is different: the user asks an LLM, the AI interprets the need, searches the internet and makes a direct recommendation. The change is radical.
In this new context, brands are no longer competing just for visibility, but to influence the algorithmic decision. Ultimately, the nature of the game has changed: you no longer compete to appear, but to be chosen.
And that is where the concept of GEO comes in: it is not about ranking higher, but about getting the AI to understand you, to consider you a reliable source… and to put you forward as the preferred option.
What factors influence a brand’s positioning in SEO and generative AI engines?
If you simplify it greatly, there are three things that are making the difference:
- Visibility, that is, the brand entering the conversation.
- Narrative, how the AI talks about you and to what extent that is faithful to what the brand wants to express.
- And authority, or rather, which sources the AI relies on to construct the response.
There is a key idea here: the AI does not rely solely on brands; it relies on third parties. Comparison sites, media outlets, forums, social media… And it is the combination of all these that ultimately determines the recommendation.
What content strategies work best for SEO and for appearing in generative AI responses?
The content that works today is the kind that helps the AI make a decision.
And that means, above all, a very different way of telling stories to what many brands are used to. AI seeks comparability, clarity and decision-making arguments. It wants to understand price versus value, differences compared to the competition, or what the product’s real attributes are.
The problem is that brands often aren’t telling that story, either because it doesn’t quite fit with their positioning or because it opens the door to uncomfortable comparisons. But if you don’t do it, your competitors will end up doing it.
If we look at a specific example, at Telefónica we can see this very clearly. Our ambition is to be the best gateway for people to access digital technologies. And in an AI environment — particularly with the shift towards generative experiences in search engines — this means that when a user asks, for example, which fibre broadband is best for them, the AI must be able to clearly understand what we offer and why we are a good choice.
That is why we are working not only on communication, but also on how that value proposition—network quality, access to the best technology, trust in the service—also appears in the environments from which AI learns: comparison sites, media or structured content that facilitates recommendations. This approach can be applied to any sector where the user’s decision depends on clear comparisons.
The challenge is to give AI what it needs to make recommendations… without compromising who you are as a brand. Because ultimately, it’s not the one with the most content that wins, but the one that aligns with the narrative AI needs to make recommendations.
How can you optimise content so that AI models (ChatGPT, Gemini, Copilot) mention your brand?
Here we do see three things that are quite clear: structure, data and authority.
First, structure. The content must be clear, modular and well-organised, so that the AI can easily extract specific ideas and reuse them in its responses.
Second, data. Not opinions or general claims, but specific, verifiable and comparable information: figures, attributes, conditions – elements that actually help build a recommendation.
And third, authority. AI doesn’t build responses solely from the brand’s website; it relies primarily on third-party sources. That’s why it’s essential that this content also exists on comparison sites, media outlets or relevant forums… and that these sources validate it.
Ultimately, optimising for AI means making your content easy to understand, easy to reuse… and ensuring it appears in the places AI trusts to build its responses.
What KPIs allow us to measure ranking in generative AI engines when there are no clicks or traffic?
For years we have measured success based on traffic and clicks. In the age of AI, that is no longer enough, because many responses no longer generate direct interaction.
That is why we need to start changing how we measure. The first step is to understand whether or not the brand is in the response, that is, whether it appears when the AI discusses your category. This relates to metrics such as share of voice in generative environments.
Then there is the role you play within that response. It is not the same to appear as just another option as it is to be the top recommendation or the option that the AI places first.
How you perform across different platforms is also becoming key, because ChatGPT, Google or Copilot do not respond in the same way, and that is where you either gain or lose visibility.
And another important point is the full journey: there are brands that are very present at the start, but disappear at the moment of decision, and that is where real impact is lost.
Ultimately, rather than measuring traffic, what we are measuring is whether the AI includes you in the decision… and, above all, at what point it does so.
How should brands adapt to SEO in the age of AI to improve their visibility on generative search engines?
I’d be cautious here, because we’re at a stage where, as an industry, we’re still working out a lot of things. But there is one thing that’s already becoming quite clear.
For me, the key is that brands need to start asking themselves what the AI is actually learning from them. Until now, you simply put out your message and that was that, but now what matters is which version of your brand ends up being retained in the models… and that doesn’t always match what you want to say.
We’re also seeing that authority is shifting even further. What others say about you — comparison sites, the media, communities — carries increasing weight, whilst what you say directly carries less. And that changes the game quite a bit.
And then there’s something I find particularly relevant: it’s not enough just to have a good reputation. Many brands are well-positioned in terms of image, but when it comes to making recommendations, they aren’t providing clear arguments. And that’s where you really win or lose.
Ultimately, I believe the change runs deeper than just a question of content. We’re moving from working on visibility to working on recommendations. And that forces you to rethink how you build brand presence across the entire ecosystem, not just on your own channels.
We’re still learning, but the direction is quite clear: whoever manages to influence how AI constructs its responses will have a very clear advantage, in a context where more than half of consumers in markets like Telefónica’s are already using AI in their decision-making process.
