Share of voice in AI
Overview
Share of voice in AI (AI SOV) measures an entity's relative presence across AI answers: its proportion of all mentions (or citations) within a defined category and prompt set, compared with competitors. It adapts the classical advertising and SEO concept of share of voice to generative and answer engines.[1]
Unlike absolute AI visibility, share of voice is inherently comparative: it situates an entity against the competitive set that the model actually surfaces for a topic.
How it is measured
AI SOV is computed over a fixed prompt set, model, and competitive set:
- AI SOV (mentions) = entity mentions ÷ total competitor-set mentions.
- AI SOV (citations) = entity citations ÷ total competitor-set citations.
Because mentions and citations diverge (see Citation rate vs Mention rate), share of voice should specify which it counts. Results are reported per model and refreshed over time, since model updates change the surfaced set.
| Term | Measures |
|---|---|
| Share of voice (AI) | Relative share vs competitors |
| AI visibility | Absolute presence of one entity |
| Mention/citation rate | Frequency, not relative share |
| Classical SOV | Share on advertising or ranked-search surfaces |
AI share of voice is not the same as an absolute visibility count: an entity can raise its mention rate while its share of voice falls if competitors rise faster.
Examples
- Across 100 category prompts, Brand X is mentioned 60 times out of 300 total competitor mentions — a 20% mention share of voice.
- The same brand's citation share of voice may be lower if it is named but rarely cited.
See also
References
- ↑ Alhena (2025). "AI Share of Voice: How to Measure LLM Brand Visibility." https://alhena.ai/blog/ai-share-of-voice-ecommerce/
