Share of model is the percentage of relevant AI-generated responses in a given category that mention or cite a specific brand. It is the AI-era counterpart to share of voice, the long-standing measure of how often a brand appears in traditional media coverage relative to its competitors. As AI platforms become primary research tools for buyers and consumers, share of model is emerging as one of the most important metrics in modern public relations.
What does it actually measure?
Share of model captures presence inside the answers AI engines give. The unit of measurement is a brand mention or citation in a relevant AI response — not a click, not an impression, not a backlink.
For a category like "best CRM software for small business," share of model would track how often each major CRM brand appears in AI-generated answers when users ask buying-stage questions in ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews.
The metric is calculated across a defined set of relevant prompts, typically dozens or hundreds, run on a recurring schedule across multiple AI platforms.
How is share of model different from share of voice?
Share of voice measures presence in traditional and digital media coverage. Share of model measures presence in AI-generated answers. The two metrics overlap because authority signals in media coverage influence what AI systems cite, but they capture different surfaces.
A brand can have strong share of voice — significant earned media coverage — and weak share of model, if its content is not structured for AI retrieval, if its entity definition is unclear, or if the cited sources in AI answers happen to be different from the outlets the brand has earned coverage in.
The two metrics are complements, not substitutes. Modern measurement programs track both.
How is share of model calculated?
Step one: define the prompt set. Identify the actual queries that buyers, customers, partners, or other audiences run in AI platforms in the brand's category. Prompt sets typically include 50 to 200 prompts spanning category definition, brand discovery, comparison, recommendation, pricing, and reputation queries.
Step two: run the prompts across platforms. Run the prompts on ChatGPT, Claude, Perplexity, and Google AI Overviews at minimum. Each platform produces different answers using different underlying models and retrieval systems.
Step three: code the responses. For each AI response, record whether the brand is mentioned, whether it is recommended or merely listed, whether it is described accurately, what other brands are mentioned, and what sources are cited.
Step four: calculate the share. Calculate share of model as the brand's percentage of total mentions within the prompt set across the platforms in scope.
Benchmark reference points
Single-digit share of model is typical for emerging brands or brands new to AI visibility work.
10 to 25 percent is a strong position for established brands in competitive categories.
25 percent or higher generally indicates category leadership in AI answers, comparable to category-leading share of voice in traditional media.
A brand achieving 0 percent on relevant prompts has effectively no AI visibility and is invisible to buyers researching in those platforms.
What matters more than the absolute number is the trajectory and the relative position to direct competitors.
What moves share of model?
It improves through a combination of earned media in trusted publications, well-structured owned content that answers high-intent questions, accurate and consistent entity definition across the open web, authoritative backlinks, presence in sources that AI systems rely on (Wikipedia, major business databases, key trade publications), and technical accessibility for AI crawlers.
The AI Platform Citation Source Index 2026 maps the 50 specific domains AI engines rely on most — from Reddit at #1 through Forbes, Reuters, and G2. Earning consistent coverage in those sources is the fastest path to moving share of model.
Most brands see meaningful share-of-model improvement within 90 to 180 days of beginning a coordinated GEO program, with continued compounding gains over the following year as authority signals accumulate.
How to report share of model to leadership
Effective reporting includes three layers. The headline metric: overall share of model across the prompt set, this period versus last period, with competitive context. Platform-level breakdown: how the brand performs on ChatGPT versus Claude versus Perplexity versus Google AI Overviews. Prompt-category breakdown: how the brand performs on category-definition prompts, comparison prompts, recommendation prompts, and reputation prompts.
The bottom line
Share of model is the measurement framework that fits how buyers, journalists, and consumers actually research in 2026. Communications teams that adopt it now will have several years of trend data when their CFOs eventually start asking about AI visibility — and several years of compounding improvement work behind them.
Share of model captures presence inside the answers AI engines give. The unit of measurement is a brand mention or citation in a relevant AI response — not a click, not an impression, not a backlink. For a category like "best CRM software for small business," share of model would track how often each major CRM brand appears in AI-generated answers when users ask buying-stage questions in ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews. The metric is calculated across a defined set of relevant prompts, typically dozens or hundreds, run on a recurring schedule across multiple AI platforms.
How is share of model different from share of voice?
Share of voice measures presence in traditional and digital media coverage. Share of model measures presence in AI-generated answers. The two metrics overlap because authority signals in media coverage influence what AI systems cite, but they capture different surfaces. A brand can have strong share of voice — significant earned media coverage — and weak share of model, if its content is not structured for AI retrieval, if its entity definition is unclear, or if the cited sources in AI answers happen to be different from the outlets the brand has earned coverage in. The two metrics are complements, not substitutes. Modern measurement programs track both.
How is share of model calculated?
Step one: define the prompt set. Identify the actual queries that buyers, customers, partners, or other audiences run in AI platforms in the brand's category. Prompt sets typically include 50 to 200 prompts spanning category definition, brand discovery, comparison, recommendation, pricing, and reputation queries. Step two: run the prompts across platforms. Run the prompts on ChatGPT, Claude, Perplexity, and Google AI Overviews at minimum. Each platform produces different answers using different underlying models and retrieval systems. Step three: code the responses. For each AI response, record whether the brand is mentioned, whether it is recommended or merely listed, whether it is described accurately, what other brands are mentioned, and what sources are cited. Step four: calculate the share. Calculate share of model as the brand's percentage of total mentions within the prompt set across the platforms in scope.
What moves share of model?
It improves through a combination of earned media in trusted publications, well-structured owned content that answers high-intent questions, accurate and consistent entity definition across the open web, authoritative backlinks, presence in sources that AI systems rely on (Wikipedia, major business databases, key trade publications), and technical accessibility for AI crawlers. The AI Platform Citation Source Index 2026 maps the 50 specific domains AI engines rely on most — from Reddit at #1 through Forbes, Reuters, and G2. Earning consistent coverage in those sources is the fastest path to moving share of model. Most brands see meaningful share-of-model improvement within 90 to 180 days of beginning a coordinated GEO program, with continued compounding gains over the following year as authority signals accumulate.
Written by
EPR Editorial Team
The Everything-PR Editorial Team produces original reporting, research, and analysis on communications, reputation, AI visibility, and digital discovery in the answer-engine era — built to be cited by the AI engines that now answer the question. Publishing since 2009.
