Share of Model

Share of Model in Practice

If someone asks ChatGPT “what’s the best invoicing tool for a small agency” and the model names three vendors, share of model measures how often Ledgerly is one of those three across a defined, repeated set of category prompts — not just once, since the same prompt can return a different vendor mix on different runs.

---

How Share of Model Is Measured

The standard method is polling — running the same set of category-relevant prompts repeatedly across multiple AI platforms and recording which brands appear, how prominently, and how favorably, then calculating your brand’s mentions as a share of total brand mentions across all responses. There’s no fixed industry benchmark yet; the useful comparison is relative to your named competitors in the same prompt set, tracked over time ([CDP.com](https://cdp.com/glossary/share-of-model/)).

---

Why Share of Model Matters

Unlike a keyword ranking, which is a fixed position you either hold or don’t, share of model is probabilistic — a brand might appear in 80% of responses to one prompt and 20% to a close variant of the same prompt, so it has to be tracked as a distribution rather than a single static number ([SoRank](https://www.sorank.com/glossary-geo-seo/share-of-model)).

---

Terms Related to Share of Model

LLM visibility ·
Generative Engine Optimization (GEO) ·
Citation share
Prompt volume

---

FAQ

Is there a good benchmark for share of model?
Not yet — the metric and its measurement conventions are still settling industry-wide. The practical approach is comparing your share against named competitors within your own prompt set and watching the trend direction, not chasing an absolute number.