Large Language Model Optimization (LLMO)
A label for optimising towards large language models. It has no documented coinage, and it collides with the established engineering meaning of optimising a model itself.
The naming collision
In machine-learning practice, optimising a large language model means making it smaller, faster or cheaper to run. Marketers adopted the same phrase for making a brand more likely to be mentioned by one.
Two established meanings for one acronym in adjacent fields is a poor outcome, and it is why the term appears here as contested rather than defined.
What it is reaching for
The underlying instinct is sound: the target is a model rather than a search index, and the mechanisms differ. But AEO and GEO already name that work, and both have documented origins.
A term with no coinage, no date and a collision with prior art is hard to defend. Using it is not wrong, but it should be glossed.