Search has a new layer.
A growing share of search journeys now end in an AI-generated answer rather than a list of blue links. ChatGPT, Gemini, Perplexity, Claude and Google's AI Overviews all synthesise information and recommend brands without the user ever reaching a results page. Generative Engine Optimisation is the work of making sure your brand turns up in those answers – accurately, credibly and consistently.
I have been working on this since before most of the industry had a name for it. And I have a fairly good idea where it is going.
Where it is going, since you asked: part of me thinks we end up back at a list of links. Once genuinely capable assistants are embedded in people's daily lives, a shortlist is a more useful answer than a paragraph. I would bet the team at Adblock have already mapped out what that looks like and where they fit into it. I can picture a small watch-sized screen showing stripped-back "results" after you have asked an assistant for three options.
AI Citation Audits
Understand how – and how often – your brand is being cited across major AI platforms. Baseline measurement before any GEO programme can be built. If we don't know where we started from, how can we measure how high we are climbing?
Learn more → GEO 02Entity & Brand Strategy
How LLMs understand your brand – your entity definition, associations, structured data and the content signals that shape what AI systems say about you.
Learn more → GEO 03LLM Visibility Tracking
Ongoing measurement of your brand's presence across AI answer engines is the objective here – tracking citation frequency, accuracy and sentiment over time as the landscape evolves.
Learn more →Common questions
- Isn't this just SEO with a new name? There is real overlap – structured data, a clear entity definition and content that answers a question rather than circling it all serve both. But the mechanism differs. Search engines rank pages; language models assemble an answer and decide who is worth citing inside it. You can rank first and still be absent from the answer, which is the part that catches teams out.
- Can you actually influence what an LLM says about a brand? Influence, yes. Control, no – and anyone claiming otherwise is overselling. What you can do is shape the source material these systems draw on: your own site, your structured data, your press coverage, the reference sources they lean on. That moves the odds considerably, and it is measurable, which is the point of starting with an audit.
- Is it too early to be spending money on this? It is early enough that the competition in most categories is thin, which is generally the good time rather than the bad one. That said, most of the groundwork is work that pays off in ordinary search too, so it is rarely a bet on one outcome. If your category has no AI answer volume worth chasing yet, I will tell you that.