AI share of voice: why it misleads

AI share of voice is the share of AI answers in a category that mention the brand, compared with everyone else mentioned. It's easy to count and easy to chart. It also hides the thing that matters.

What it counts

Every time the brand's name appears in an answer, it counts. It doesn't matter whether the engine:

• put the brand first and called it the best,
• listed it ninth of ten with no comment, or
• named it to warn buyers away.

All three add the same to share of voice.

Why a recommendation beats a mention

A buyer who asks an AI engine "which should I choose?" usually acts on the first one or two names it backs. Being in the list isn't being chosen. A brand with a high share of voice can still be the one the engine steers people away from.

The verdict-first alternative

Read every answer as a verdict:

• Top pick: the brand is named first, or highlighted as the best option available.
• Recommended: the engine suggests the brand as a strong, reliable choice.
• Mentioned in passing: the name appears, but the engine doesn't back it.
• Named with caution: the engine names the brand but warns buyers about it.
• Not mentioned: the brand doesn't show up at all.

Then ask: on how many engines is the brand recommended, and what sentence proves it? Numbers support the verdict; they don't replace it.

Keep Found and Recommended apart

• Found counts only questions that leave the brand name out.
• Recommended can also come from questions that name the brand.

Mixing them inflates the result: of course an engine talks about a brand when the question names it.

See it on the brand

Get your Free Report for the verdict and the quote on six engines, or read the Seenford Standard.