Share of Voice (AI Search)

Share of voice (SOV) translates a classic media metric to generative answers. Instead of measuring ad impressions or organic rankings, it measures the slice of AI recommendations your brand owns relative to competitors. The formula is straightforward: your brand's mentions divided by total mentions of all tracked brands across the same prompts and engines, over the same period.

SOV is inherently competitive, which makes it more informative than absolute counts. A mention rate of 40 percent sounds strong until you learn a competitor sits at 80 percent on the same prompts. SOV exposes that gap directly.

AI assistants compress shortlists: where a buyer once reviewed ten vendors, the assistant proposes three. The brands inside that shortlist split the demand, so SOV in AI answers maps closely to share of consideration in the category. Marketing research has long linked share of voice to market share growth, and the relationship tightens when the medium itself, the AI answer, performs the recommending.

SOV also reveals engine-level asymmetries. A brand can hold 50 percent SOV on Perplexity but 10 percent on Gemini, because each engine retrieves different sources. Tracking SOV per engine via competitive benchmarking shows exactly where to invest.

Define the competitor set and the prompt set first; both choices determine the denominator. Then collect answers on a fixed schedule, extract every brand mention, and compute each brand's percentage. Watch trends after model updates, competitor launches, and your own content pushes to attribute movement.

Growing SOV means winning prompts you currently lose: study the sources cited in answers that favor competitors, strengthen or place content there, and re-measure. Geonimo's competitor tracking computes mention-share across all tracked rivals daily, turning SOV into an operational dashboard rather than a quarterly study.