METRICS

Share of Voice in AI Answers

A dense particle cloud with varied sub-clusters, where one sub-cluster's density dominates the total mass, representing a brand's mention proportion among all brands' total mentions
Share of Voice is the proportion of a brand's mentions relative to the total mentions of all brands in the same set of questions

IN ONE SENTENCE

Share of voice is a brand's mentions as a proportion of all brand mentions across the same question set.

Share of voice is a brand's mentions as a proportion of all brand mentions across the same question set.

OUR POSITION

It answers a different question from mention rate: mention rate asks whether you appear, share of voice asks how much of the conversation is yours. In crowded categories the second is more diagnostic.

01

How it is calculated

Numerator is the brand's mentions; denominator is all brand mentions across the same batch of answers. Answers routinely name several brands, so the denominator is usually larger than the answer count.

Normalise brand names first — variants, abbreviations and parent-company names have to be merged, or the share gets split across spellings.

02

The step that goes wrong most often

⚠️ The brand list has to be fixed before measuring. Miss a competitor and your share is overstated; include an unrelated same-name term and it is understated.

Automated brand extraction has a high false-positive rate — tables inside answers run text together, so both word boundaries leak. Anything reported externally needs manual review.

03

How to read it

Read the trend, not the absolute value. Changing the brand list invalidates the history, so the meaning lives in movement under one fixed definition.

Read it per platform. The brand mix for the same questions varies sharply across platforms, and a blended figure hides a decline on one of them.

Data behind this page

40.1%

Reddit's share of citations in AI answers

SourceSemrush, 150,000+ citations across 5,000 keywords,2025-06

under 5%

Ceiling on any single domain's citation share on one platform

SourceEvertune, 200M prompts,2026

Sources

  1. [1]GEO: Generative Engine Optimization.Aggarwal et al., KDD 2024.2024

Updated 2026-08-10