What Is AI Share of Voice?

AI share of voice (AI SoV)is the percentage of mentions your brand gets across AI answer engines — ChatGPT, Perplexity, Claude, Google AI Overviews — for a fixed set of prompts, relative to your competitors. The formula: AI SoV = (your brand's mentions across a prompt set) ÷ (total brand mentions across that same prompt set, all competitors included). Run the same 50–500 prompts every month, count who gets named, cited or recommended, and you have a repeatable, comparable number — the AI equivalent of traditional share of voice.

The concept borrows directly from advertising and PR, where share of voice has long meant a brand's portion of the total conversation in its category — how much of the ad spend, media mentions or social conversation belongs to you versus rivals (see Sprout Social's definition, linked below). AI SoV applies the same logic to a new channel: instead of counting impressions or backlinks, you count how often a large language model volunteers your name when someone asks it a category question like "best project management software" or "top CRM for startups."

Why it matters now: AI answer engines are becoming a real discovery surface, and they don't rank websites the way Google's ten blue links do — they synthesize an answer from whichever sources they trust and cite a handful of them. In our own sample of 500 prompts and 2000 model calls, the average AI answer cited 9.6 sources, and 84% of those citations pointed to user-generated content (forums, reviews, Reddit threads) rather than brand or publisher websites. That is a fundamentally different citation pattern than classic SEO, which is why tracking AI SoV separately from search rankings matters — a page that ranks #1 on Google can still be invisible to an AI answer, and vice versa. Google itself documents how AI Overviews generate summaries from web content rather than simply surfacing the top organic result (see the AI Overviews documentation, linked below).

A brand with high AI SoV shows up consistently across a prompt set — not just once, by luck, but as a repeatable pattern the model reaches for. A brand with low AI SoV might rank well in Google yet be absent from the sources an AI answer actually cites: in our benchmark, 60% of AI-cited sources sat outside the top 10 Google organic results for the same query, meaning classic rank tracking alone would miss most of what AI engines are actually reading.

How AI SoV differs from traditional share of voice

Traditional share of voice is usually computed from ad spend, media mentions or social listening data — inputs you can buy or measure at scale. AI SoV instead requires running actual prompts against actual models and reading the actual answers, because there is no public API that reports "how often ChatGPT recommends brand X." That makes the prompt set itself the unit of measurement: choose prompts real buyers would ask, run them repeatedly, and track mentions, citations and recommendation strength over time.

How to read this index

The AI Share of Voice Index tracks these mentions and citations across verticals so brands can see where they stand relative to competitors, and which sources AI engines actually trust in their category — without having to run the prompt testing themselves.