Definition
LLMO (Large Language Model Optimization) is an umbrella term for influencing how language models represent a brand, covering both what a model absorbed during training and what it retrieves at answer time. In practice most published LLMO advice is about the retrieval half, because that is the half anyone can affect this quarter.
Why it matters
The distinction decides what is worth doing. Training-time presence is slow and largely out of your hands; retrieval-time presence responds to pages published this month and to third-party sources the model reaches for.
How to measure it
Ask the same question with and without web search enabled where the tooling allows it. A brand that appears only with search on is living entirely in the retrieval half.