Skip to main content
Measurement · Emerging

​
Definition

LLM brand recall is the accuracy and completeness with which a specific large language model can reproduce correct information about a brand from its parametric knowledge — without retrieval augmentation. It measures what the model “knows” about a brand from training data alone. LLM brand recall is the training data dimension of AI search visibility. A brand with strong LLM recall is represented accurately in model weights — the model can correctly state what the brand does, who it serves, where it operates, and what distinguishes it, without needing to retrieve external sources. Strong recall is built through training corpus presence: being accurately described in Wikipedia, Wikidata, widely-referenced web content, and authoritative publications before a model’s training cutoff. Recall degrades over time as models age relative to their cutoffs, which is why monitoring recall across model generations is part of a complete AI search management program.

LLM brand memory

Training corpus

Knowledge cutoff

Brand grounding

Prompted citation

​
Relevant PLC Services

AI Search Visibility Assessment Context Map Entity SEO