Share of model
How often, and how favorably, AI assistants name your brand across a defined set of buyer questions, measured against competitors.
In depth
What it really means
Share of model counts mentions, linked or not. It answers a blunt question: when a buyer asks an assistant for options in your category, how often is your name in the answer, and how often is a competitor’s?
It is the closest thing AI search has to share of voice, and it captures something citation share misses. You can be recommended by name in an answer that cites nobody. For consideration-stage queries, the mention is the outcome that matters.
How it works
- Define a fixed set of buyer questions.
- Run each one repeatedly across each assistant.
- Log every brand named, plus the qualifier attached to it.
- Compute your share of total mentions and track the trend.
Share of model = (mentions of your brand across the query set / total brand mentions across the query set) x 100
Weight by sentiment, or the number lies. A negative mention should not count the same as a recommendation. Run each query at least three times per assistant and average. Track per assistant, never blended.
Pros & cons
Pros
- It measures the outcome you actually want, being recommended.
- It works on queries where no sources get cited at all.
- Competitive comparison is direct and easy to explain to leadership.
Cons
- Manual measurement is slow, and tools are expensive.
- Non-determinism means small changes are indistinguishable from noise.
- Mentions do not attribute to a page, so you cannot tell what earned them.
Common mistakes
- Counting mentions without recording sentiment.
- Blending assistants into one number when they behave very differently.
- Changing the query set between periods, which breaks the trend.
- Measuring brand-name queries, where you always appear, instead of category queries where the competition is.
Best practices
- Build a query set of category and comparison questions, not brand-name questions.
- Freeze the set, then run it monthly across all four major assistants.
- Record the exact qualifier used alongside each mention.
- Compare only against your real competitive set.
- Pair it with citation share, since the two move independently.
FAQs
What is share of model?
How often and how favorably AI assistants mention your brand across a fixed set of buyer questions, compared with competitors. It is the AI equivalent of share of voice.
How do I measure share of model?
Freeze a set of category questions, run each several times across each assistant on a schedule, and log every brand named plus the qualifier attached to it.
How is it different from citation share?
Share of model counts brand mentions. Citation share counts linked sources. You can be recommended by name in an answer that cites nobody, which citation share would score as zero.
What is a good share of model?
It only means something relative to your real competitors, and the trend matters more than the level. Rising share on category queries is the signal worth reporting.
Keep reading
Related on LymLyt
Beyond LymLyt
Further reading
Want this working on your site?
We build the content behind the term, ranked in search and cited by AI.
Book a 30-min call →