I'm the EMEA Marketing Director at Bang & Olufsen. Over the past year I've been running a simple experiment: asking ChatGPT, Perplexity, and Claude for brand recommendations in our category — premium audio — and tracking which brands appear, how often, and with what level of specificity.

The results were uncomfortable. Some brands consistently surface with rich, accurate descriptions. Others — including ours at times — either don't appear or appear with vague, outdated framing. This has nothing to do with SEO. It's about how well a brand's identity is structured in the data that LLMs train and retrieve from.

I've been calling this GEO — Generative Echo Optimisation. The core idea: AI doesn't find your brand, it reflects it. If your brand's presence online is fragmented, inconsistent, or lacks depth, the AI produces a faint echo or none at all.

A few things I've been trying to measure: - Share of Echo Voice (SEV): % of times your brand appears vs. competitors in relevant AI recommendation queries - Brand Accuracy Score (BAS): how closely AI descriptions match your intended positioning - Conversational Depth Index (CDI): how much detail AI generates about your brand vs. surface-level mentions

The problem is there's no standard tooling for this yet. I'm doing it manually with prompt templates and spreadsheets.

Has anyone else been tracking this? Are there tools I'm missing? And do you think GEO is a real discipline or just repackaged content strategy with an AI label?

(I wrote a short book on this if anyone wants the framework — but genuinely more interested in what others have found.)