Share of Model: The New Metric That Replaces Share of Voice

For decades, "share of voice" has been the standard metric for measuring brand visibility in advertising and media. It tells you what percentage of the total conversation in your market belongs to your brand. Now, a new metric is emerging that applies the same logic to AI-powered search: Share of Model. This metric measures how frequently your brand is cited, recommended, or mentioned by AI models relative to your competitors, and it may be the single most important metric for digital brand visibility in the coming years.

This article explains what Share of Model is, why it matters, how it is calculated, and what you can do to improve yours.

Defining Share of Model

Share of Model (SoM) is a metric that quantifies your brand's presence in AI-generated responses as a proportion of all brand mentions within your competitive category. If you and four competitors are all relevant to a particular query, and AI models cite your brand in 40% of their responses to that query, your Share of Model is 40%.

The metric can be calculated at various levels of granularity: for specific queries, for query categories, for individual AI platforms, or as an aggregate across all platforms. Each level provides different strategic insights.

Why Share of Model Matters

In an era where AI search is growing rapidly and zero-click searches are the norm, traditional metrics like organic traffic and keyword rankings tell an increasingly incomplete story. A brand might rank first on Google for a key term but be completely absent from AI-generated responses to the same query. Conversely, a brand with modest traditional rankings might be the top-cited result across AI platforms.

Share of Model fills this measurement gap. It provides a direct measure of your brand's influence in the channels where a growing proportion of your audience discovers information, makes decisions, and forms brand preferences. As AI search adoption continues to accelerate, SoM will become as fundamental to marketing measurement as share of voice has been for traditional media.

Share of Model is not just a vanity metric. It is a leading indicator of future market share. The brands that AI models recommend today are the brands that consumers will choose tomorrow. Measuring and improving your SoM is not optional; it is strategic imperative.

Aether Insights, 2026

How Share of Model Is Calculated

Calculating Share of Model requires systematic querying of AI platforms and careful analysis of their responses. Here is the methodology.

Step 1: Define Your Query Set

Identify the queries that matter most to your business. These should include branded queries (your brand name), category queries (your service or product category), comparison queries (queries that pit competitors against each other), and informational queries (questions about topics in your area of expertise). A typical audit might include 50 to 200 queries depending on the complexity of your business.

Step 2: Query Multiple Platforms

Run each query across every major AI platform: ChatGPT, Perplexity, Google AI Overviews, and Claude. Record the full response for each query on each platform. Because AI responses can vary between sessions, run each query multiple times over several days to capture the range of responses.

Step 3: Analyse Brand Mentions

For each response, identify every brand mention. Categorise mentions as primary (your brand is a main recommendation), secondary (your brand is mentioned as an alternative or comparison), or contextual (your brand is mentioned in passing). Weight these categories differently in your SoM calculation, with primary mentions carrying the most weight.

Step 4: Calculate the Metric

For each query, divide your weighted brand mentions by the total weighted mentions of all brands in your competitive set. Aggregate across queries and platforms to produce your overall Share of Model score. Track this metric over time to measure the impact of your GEO efforts.

1. Automate where possible: Manual querying is time-intensive and inconsistent. Tools like Aether AI can automate the querying, recording, and analysis process, providing consistent, repeatable measurements.
2. Track competitors consistently: Your SoM is relative, so you must track competitor mentions with the same rigour as your own. Changes in competitor SoM can be as strategically important as changes in your own.
3. Segment by platform: Your SoM may vary significantly across platforms. Understanding these variations helps you target your optimisation efforts where they will have the greatest impact.
4. Segment by query type: Branded queries will naturally produce high SoM scores. Focus your competitive analysis on category and informational queries, where the real competitive dynamics play out.

Strategies to Improve Your Share of Model

Improving your Share of Model requires a comprehensive approach that addresses every factor influencing AI citation decisions. The following strategies have proven most effective.

• Strengthen your entity definition: Ensure your brand is clearly defined through comprehensive schema markup, consistent NAP data, a well-maintained Wikipedia presence, and accurate directory listings. A clearly defined entity is easier for AI models to cite confidently.
• Produce authoritative, citable content: Create content that AI models want to reference, including original research, definitive guides, expert analyses, and detailed case studies. Every piece of high-quality content is a potential citation source that increases your SoM.
• Build off-site authority: Invest in digital PR, thought leadership, and industry contributions that generate brand mentions across authoritative external platforms. Off-site content is a powerful driver of AI citation confidence.
• Optimise technical infrastructure: Implement llms.txt, ensure AI crawler access, and maintain comprehensive structured data. Technical excellence removes barriers between your content and AI citation.
• Monitor and respond to AI representations: Regularly check how AI models describe your brand and correct any inaccuracies through content updates, schema adjustments, and off-site consistency improvements.

The Competitive Dynamics of Share of Model

Share of Model has an important characteristic that distinguishes it from many other marketing metrics: it is a zero-sum game within any given query. When your SoM increases, your competitors' SoM necessarily decreases, and vice versa. This creates strong competitive dynamics where first-mover advantages can be significant and sustained.

Brands that establish strong SoM early benefit from a compounding effect. As AI models learn to associate your brand with particular topics and services, they cite you more frequently, which generates more engagement signals, which further reinforces your position. Breaking into a category where a competitor has established dominant SoM is significantly harder than establishing leadership in a category where no brand has yet invested in AI visibility.

Integrating SoM into Your Marketing Dashboard

Share of Model should not exist in isolation. It is most valuable when integrated into your broader marketing measurement framework alongside traditional SEO metrics, brand awareness data, and business performance indicators. Track how changes in SoM correlate with branded search volume, direct traffic, lead quality, and conversion rates to build a complete picture of the business impact of your AI visibility efforts.

The transition from measuring clicks to measuring AI influence is one of the most significant shifts in marketing measurement in a generation. Share of Model provides the framework for making this transition systematically and strategically. Start measuring it now, and you will have the data foundation to guide your brand through the AI search revolution with confidence and precision.

Share of Model by Industry: Benchmarks and Expectations

Share of Model benchmarks vary significantly across industries, influenced by the number of competitors, the maturity of the market's AI presence, and the specificity of user queries. Understanding industry-specific benchmarks helps brands set realistic targets and measure progress against relevant standards.

What is a good Share of Model score for my industry?

In fragmented markets with many competitors, such as digital marketing agencies or SaaS tools, a category-leading SoM of 15-25% on non-branded queries is considered strong. In concentrated markets with fewer major players, such as enterprise cloud computing or commercial aviation, leading brands may achieve SoM scores of 40-60%. For highly local markets, such as regional service providers, SoM scores can reach 50-70% for geo-specific queries. The critical insight is that absolute SoM numbers matter less than your trajectory and your position relative to competitors. A brand that is improving its SoM by 2-3 percentage points per quarter is on a strong growth trajectory regardless of its current absolute score.

The variance across AI platforms also matters. A brand might have a 30% SoM on Perplexity (which tends to favour niche, authoritative sources) but only a 10% SoM on ChatGPT (which draws from a broader pool of well-known brands). Understanding these platform-level variations allows you to target your optimisation efforts where the greatest improvement opportunities exist.

How does Share of Model correlate with business outcomes?

Early data suggests a meaningful correlation between Share of Model and downstream business metrics. Brands that achieve and maintain category-leading SoM scores report increased branded search volume, higher direct website traffic, improved conversion rates among visitors who arrive having already encountered the brand in AI responses, and stronger performance in competitive pitches and RFP processes where decision-makers increasingly use AI tools for vendor research. While causation is difficult to establish definitively, the correlation is strong enough that leading marketing teams now include SoM as a key performance indicator alongside traditional metrics.

"Share of Model is to AI search what share of voice was to broadcast media: the metric that separates brands that are shaping the conversation from those that are invisible within it. Every marketing dashboard should include SoM alongside organic traffic, conversion rate, and customer acquisition cost."

— Tom Critchlow, Strategy Consultant and Author of "The Strategic Independent"

Common Share of Model Measurement Pitfalls

As SoM is still an emerging metric, many brands fall into measurement traps that produce misleading results. Being aware of these pitfalls ensures your SoM data is reliable and actionable.

• Over-reliance on branded queries: Your SoM for branded queries (where someone searches your name directly) will naturally be high and is not representative of competitive dynamics. Always weight your SoM analysis toward category and informational queries where real competitive positioning is determined.
• Insufficient query volume: Running a handful of queries once provides a snapshot, not a trend. Reliable SoM measurement requires a consistent set of at least 50 queries, run across all major platforms, repeated at regular intervals over multiple months.
• Ignoring response variability: AI models can produce different responses to the same query on different occasions. Run each query multiple times over several days and average the results to account for this variability.
• Conflating mentions with recommendations: Not all brand mentions are equal. Being named as the primary recommendation carries far more weight than a passing contextual reference. Your SoM methodology should weight mention types accordingly.