AI reshapes your brand based on what consumers think of you — not what you aspire to be through owned media.
In traditional marketing, you survey users to understand how they see you. You measure every month or six months to track what’s changed. You have some control mechanisms — FMCG brands spend a lot of time shaping perceptions.
I’ve been building category entry point analysis across AI models for the last few months. What it reveals is fascinating. You can map exactly how AI organises a brand across the key moments that matter to the consumer.
Take fast casual restaurants as an example (see the CEP table above):
- When people eat there.
- Why they choose it.
- Who they go with.
What we don’t understand yet is whether the act of prompting AI for products and services is itself shaping how people perceive brands.
AI reads your signal environment — customer reviews, Reddit threads, comparison articles, journalist shorthand.
The inputs that shape AI brand perception are overwhelmingly earned, not owned. The model assembles a version of your brand from whatever it finds, and that version has its own architecture.
In traditional media, the brand controls the message. In AI, the medium has its own logic for assembling meaning. You can influence the inputs, but the model decides what to build with them.
That requires understanding the gap between your intended brand architecture and the one AI has already built for you.
That’s what I’m building with ATTENT10N — a brand performance platform for AI, helping us understand how AI understands, ranks, positions and compares brands. It’s a completely different playing field, where old rules and new rules come together just as media and interfaces are merging.
If this is useful, subscribe below for more.