From SEO to GEO: brands need to be named by AI
As purchase decisions increasingly pass through AI assistants, the question is no longer your Google rank but whether AI chooses to mention you.
From ranking to recommendation
Kantar predicts that purchase decisions will increasingly be mediated by generative AI and agents that recommend brands and content. Consultants call the practice GEO (Generative Engine Optimisation) or AEO. The difference: the goal is to be chosen by AI as a trusted source to mention, not just to sit at some position in a results page.
If the model doesn't know you, it won't choose you.
GEO is a brand problem
For AI to understand who you are and what you offer, content needs to be clear, structured and machine-readable. A few recommendations repeat across sources:
- Answer first, explain after: models tend to pull from the opening lines under a heading.
- Quotable content: short answers with supporting evidence.
- Trust signals: real reviews, expert partners, active communities, consistent facts across platforms.
Differentiation is still the advantage
Kantar warns that salience alone is not enough: brands that fail to differentiate get lost in a sea of sameness. Positioning remains the foundation, and GEO is how you put that positioning into a structure both people and machines can read. And technology doesn't buy things; people do.
What to do now
- Rewrite service pages and FAQ as questions and answers, answering in the first sentence.
- Add structured data (schema) and an AI-facing description file such as llms.txt.
- Keep brand facts identical across the website, social and business profiles.
- Monitor what AI says about you, then fix the sources that mislead it.
A note: many visibility-uplift figures come from GEO service vendors and are self-reported; treat them as direction, not promises.