← Glossary

What is AI share of voice?

AI share of voice is how often AI answers name a brand compared with its competitors, measured across a fixed set of buyer questions asked in tools such as ChatGPT, Perplexity and Google's AI Mode.

Cited from the66th.com/glossary/ai-share-of-voice

Liam Lytton Founder, The 66th Published 3 November 2026Updated 3 November 2026

How AI share of voice works

You write down the questions your buyers ask, from first research to ready to buy. You ask each AI tool every question, record every business each answer names, and count. A brand's share is its mentions divided by all the brand mentions across the answers.

The number means something only when the questions stay the same, so each month compares with the last.

BusinessAnswers that name it (of 20)Share of all mentions
The client420%
Competitor A945%
Competitor B735%
Illustrative. 20 buyer questions for a plumber in Calgary, asked once in ChatGPT. Business names are made up.

Why you have to measure it yourself

Google's Generative AI performance reports in Search Console show how often URLs from your own site appeared in AI features on Google Search and Discover. They show nothing about competitors, and nothing about answers that name you without linking to you.

Answers also differ by tool. Google's guide to AI features says even its own AI Overviews and AI Mode may use different models and techniques, so the responses and links they show vary. Share of voice has to be counted tool by tool.

What it means for your business

AI share of voice shows who the AI tools put in front of your buyers when you are not named. A low share on questions asked close to purchase shows which competitors the tools name to those buyers instead. It matters for services people research and compare before they call.

It is a comparison, so read it with the absolute counts. A share can rise because a competitor dropped out of answers while your own mentions stayed flat. Raising your own count is the work of our generative engine optimization service.

Common mistakes

  • Changing the questions between months. The share then measures the new questions, and the trend is lost.
  • Measuring one tool and calling it AI share of voice. Google says its own two AI features can answer differently, so count each tool.
  • Asking questions that already contain your brand name. Those answers name you because the question did; use the questions buyers ask before they know you.
  • Counting every mention as a win. Record whether the answer recommends you, lists you with a caveat, or names you as an option to avoid.

Sources: Google Search Central Blog, Hillel Maoz and Moshe Samet, Introducing Search Generative AI performance reports in Search Console, 3 June 2026; Google Search Central, AI features and your website, 10 December 2025.

Be the one customers find when they go looking.

Book a call →