Share of Model Benchmarking: How to Measure Your AI Visibility Against Competitors

In traditional marketing, Share of Voice tells you how visible your brand is relative to competitors across media channels. In the age of AI search, a new metric is needed: Share of Model, or SoM. Share of Model measures the percentage of AI-generated responses in your category that mention or cite your brand compared to your competitors. It is the single most important metric for understanding your competitive position in AI search, yet only 7% of UK businesses currently track it (Industry Survey 2026). The brands that measure SoM gain a strategic advantage that compounds over time, because what gets measured gets improved.

This guide defines Share of Model precisely, explains how to calculate it across multiple AI engines, provides benchmarks for evaluating your performance against industry standards, and shows how to translate SoM data into strategic decisions that drive measurable improvements in AI visibility. The methodology is drawn from Aether Research across hundreds of brands and the practical experience of implementing SoM tracking for clients across diverse industries.

What Share of Model Means and Why It Matters

Share of Model is the percentage of AI-generated responses for a defined set of industry-relevant queries that cite or mention your brand. If you monitor 200 relevant queries across AI engines and your brand appears in 40 of the responses, your SoM is 20%. If a competitor appears in 68 of those same responses, their SoM is 34%. This simple ratio provides an immediate, actionable picture of your competitive position in AI search.

SoM matters because it directly predicts business outcomes. Aether Client Data shows that SoM improvements of 10 percentage points or more correlate with a 28% increase in AI-attributed enquiries. This correlation is not coincidental. When AI engines cite your brand more frequently, more potential customers encounter your name in trusted, authoritative contexts. Each citation is effectively a recommendation from a source the user has chosen to consult. Higher SoM means more recommendations, which means more enquiries.

SoM Versus Traditional Share of Voice

Share of Model differs from traditional Share of Voice in several critical ways. Share of Voice measures presence across paid and organic media where your brand actively participates. Share of Model measures presence in AI-generated responses where your brand is selected by an algorithm based on content quality, authority, and relevance. You cannot buy SoM through advertising spend. You earn it through content strategy, entity authority, and sustained investment in AI-visible content.

This distinction makes SoM both harder to influence and more valuable once earned. A competitor cannot simply increase their budget to displace you from AI citations. They must produce genuinely better, more authoritative, more recent content. Conversely, the SoM you build through sustained content investment is more durable than paid media visibility because it is earned through demonstrated expertise rather than purchased through media spend.

The Competitive Intelligence Dimension

SoM data reveals not just your own performance but your competitors' strategies. When you track SoM across your competitive set, you can identify which competitors are investing in AI visibility, which topics they are winning, which engines favour them, and where their strategies have gaps. This intelligence informs both defensive strategy (protecting your existing citations) and offensive strategy (targeting topics and engines where competitors are weak). For real-time competitive monitoring, see our guide on competitor citation intelligence.

"Share of Model is the north star metric of the AI search era. It tells you not just whether your content is good, but whether it is good enough to be recommended by the systems that increasingly mediate how customers discover and evaluate businesses. If you are not measuring it, you are navigating blind."

— Rand Fishkin, Co-Founder, SparkToro

How to Calculate Your SoM

Calculating Share of Model requires a systematic approach that produces reliable, comparable data over time. The process involves three phases: query set definition, multi-engine monitoring, and SoM calculation.

Phase 1: Defining Your Query Set

The accuracy of your SoM measurement depends entirely on the quality of your query set. The set should include between 100 and 500 queries that reflect the actual questions your target audience asks AI engines. These queries should span four categories: informational queries ("What is the best approach to X?"), recommendation queries ("Which companies provide X?"), comparative queries ("How does X compare to Y?"), and geo-specific queries ("Who provides X in Y region?").

Each query should be validated against actual search behaviour data to ensure it represents genuine user demand. Avoid including vanity queries that mention your brand by name, as these inflate your SoM without reflecting organic discovery. The query set should also include queries where you currently have no visibility, ensuring that your SoM calculation captures both your successes and your blind spots.

The query set should be reviewed and updated quarterly. As your industry evolves, new questions emerge and old ones decline in relevance. A stale query set produces misleading SoM figures that do not reflect current market dynamics.

Phase 2: Multi-Engine Monitoring

Each query in your set should be submitted to all major AI engines: ChatGPT, Perplexity, Google AI Overviews, Claude, Microsoft Copilot, and Gemini. For each response, the monitoring system records whether your brand is mentioned, whether competitors are mentioned, the prominence of each mention (primary recommendation, one of several, or passing reference), and the sentiment of the mention.

Monitoring should occur at least weekly to capture the dynamic nature of AI responses. A single measurement provides a snapshot, but weekly measurements reveal trends. Is your SoM growing, stable, or declining? Are specific competitors gaining ground? Are there seasonal patterns in citation frequency? These questions can only be answered through consistent, repeated measurement. For practical guidance on implementing the monitoring infrastructure, see our article on real-time citation tracking.

Phase 3: Calculating the Score

The basic SoM formula is straightforward: SoM = (Your brand citations / Total brand citations) x 100. If your query set generates 300 responses that contain at least one brand citation, and your brand appears in 75 of those responses, your SoM is 25%.

For more nuanced analysis, calculate weighted SoM that accounts for citation prominence. A primary recommendation carries more value than a passing mention. A common weighting scheme assigns 1.0 to primary recommendations, 0.5 to listings within a group of recommendations, and 0.2 to passing mentions. Weighted SoM provides a more accurate picture of your actual visibility impact compared to the raw percentage.

Calculate SoM at three levels: aggregate SoM across all engines and queries, engine-specific SoM for each AI platform, and topic-specific SoM for each major topic cluster in your query set. This multi-level analysis reveals whether your visibility is broad or concentrated, and whether your strengths align with the topics and engines that matter most for your business.

Benchmarking Against Industry Standards

Raw SoM figures are most meaningful when compared against industry benchmarks. The competitive context determines whether a given SoM represents strong performance or untapped potential.

Category Leader Benchmarks

Aether Research (2026) data across multiple industries establishes that the average SoM for category leaders is 34% across AI engines. This figure varies significantly by sector. In professional services, where fewer brands compete for relatively niche queries, category leaders may achieve SoM above 45%. In broad consumer categories with many competitors, category leaders typically hold SoM between 20% and 30%.

These benchmarks serve as aspirational targets for brands beginning their SoM tracking journey. However, the more immediately actionable benchmark is your SoM relative to your direct competitors. If the category leader holds 34% SoM but your three nearest competitors hold 12%, 15%, and 18%, your strategic priority is different than if those competitors hold 28%, 30%, and 32%.

Engine-Specific Benchmarks

SoM varies significantly across AI engines, and understanding these variations is essential for strategic prioritisation. Perplexity tends to produce the highest concentration of brand citations because its retrieval-augmented architecture explicitly surfaces and attributes sources. ChatGPT citations are less frequent but carry significant weight due to the platform's massive user base. Google AI Overviews citations are emerging as particularly valuable because they appear within the world's most-used search engine.

Track your SoM for each engine independently. A brand might hold 30% SoM on Perplexity but only 8% on ChatGPT, suggesting that their content is well-structured for retrieval-augmented systems but lacks the training-data presence needed for ChatGPT citations. This diagnosis enables targeted strategy adjustments for each engine. For a broader framework on which metrics matter most, see our article on GEO reporting metrics that matter.

Temporal Benchmarks

In addition to competitive benchmarks, track your SoM against your own historical performance. Month-over-month and quarter-over-quarter comparisons reveal whether your strategy is producing improvements. Aether client data indicates that well-executed GEO strategies typically produce SoM improvements of 3 to 5 percentage points per quarter in the first year, with gains accelerating as content authority compounds. If your SoM is stagnant or declining despite active content investment, this signals a strategic problem that requires investigation.

Using SoM Data to Drive Strategy

SoM data is only valuable when it informs strategic decisions. The most effective implementations translate SoM analysis into specific, prioritised actions that drive measurable improvements.

Gap Analysis: Where Are You Losing?

Begin with a gap analysis that identifies the queries and topics where competitors hold citations and you do not. These gaps represent your highest-priority content opportunities. For each gap, assess why you are absent: is it a content coverage gap (you have not published on this topic), a content quality gap (your content exists but is not competitive), or a freshness gap (your content is outdated compared to competitors)?

Each gap type demands a different response. Coverage gaps require new content creation. Quality gaps require substantial improvement of existing content, including better structure, stronger sources, and more authoritative analysis. Freshness gaps require updating existing content with current data and republishing with updated timestamps. Prioritise gaps by business value, targeting the queries most likely to drive enquiries before those with lower commercial intent.

Strength Protection: Where Are You Winning?

SoM analysis also identifies your citation strengths, the topics and queries where you hold a leading position. These strengths require protection. Monitor competitor activity in your strongest areas and proactively update your content before competitors can displace you. Publish supporting content that deepens your topical authority in areas where you lead, making it harder for competitors to challenge your position.

Strength protection is often overlooked in favour of gap-closing, but it is equally important. Losing an established citation is more costly than failing to gain a new one, because re-earning a lost citation requires overcoming the momentum of the content that displaced you. Proactive maintenance of your strongest positions is therefore one of the highest-return activities in GEO strategy. For more on tracking competitor movements in your strongest areas, see our article on citation velocity tracking.

Resource Allocation: Where Should You Invest?

SoM data enables evidence-based resource allocation. Rather than spreading content investment evenly across all topics, concentrate resources on the areas where SoM data indicates the greatest return potential. Topics with high business value, active competitor investment, and achievable gap sizes should receive priority investment. Topics with low business value or insurmountable competitive advantages held by dominant players may warrant reduced investment.

Review resource allocation quarterly based on updated SoM data. As your strategy produces results, the competitive landscape shifts. Topics that were high-priority gaps may become established strengths. New gaps may emerge as competitors adjust their strategies. Quarterly re-allocation ensures your resources remain directed at the opportunities with the greatest current potential.

"Most businesses pour resources into content creation without any visibility into whether that content is actually being recommended by AI engines. Share of Model gives you the feedback loop that turns content investment from an act of faith into a data-driven discipline. It is the metric that makes GEO strategy accountable."

— Aether Insights, 2026