What is share of model?
Share of model is the frequency with which a brand is mentioned or cited by large language models when answering category-relevant questions, expressed as a percentage relative to the total mentions of all competitors in the same set. It is the AI-era equivalent of share of voice in traditional brand measurement.
The metric emerged alongside AEO and LLM visibility tracking. As AI search becomes a primary research channel for buyers, the brands that appear most consistently in AI-generated answers for category queries gain consideration that never registers in Google Search Console. Share of model captures this new form of visibility.
How to measure and improve share of model
Measuring share of model requires a standardized prompt set: the category questions your target buyers ask AI engines. Run those prompts across platforms (ChatGPT, Perplexity, Gemini, Claude) on a fixed schedule, record which brands are mentioned for each, and calculate each brand’s share of total mentions across the full prompt set.
Improving share of model follows the same logic as AEO and LLMO: build authoritative third-party mentions, maintain consistent entity descriptions, publish citation-ready content, and ensure technical accessibility. Brands with the deepest, most accurate representation across training and retrieval corpora dominate model mentions without buying a single ad.
Example
Example
An agency runs 20 prompts related to "B2B SaaS SEO agencies" across ChatGPT, Perplexity, and Gemini. Brand A appears in 14 of 20 prompts (70%), Brand B in 8 (40%), Brand C in 3 (15%). Brand A has the highest share of model and owns the AI-search consideration layer for this category.
Frequently asked questions
Is share of model the same as LLM visibility?
LLM visibility typically refers to whether a brand appears at all in model outputs. Share of model is a comparative metric: how often your brand appears relative to your competitive set. Both track AI-era brand presence; share of model makes it competitive.
How often should share of model be tracked?
Monthly is the recommended baseline. AI engine retrieval indexes and model updates can shift citation patterns quickly, so monthly tracking catches movements that quarterly reviews miss.