What Is Share of Model Voice?

Search visibility used to be easier to measure.

You checked rankings.

You checked impressions.

You checked clicks.

You checked traffic.

Those metrics still matter, but they no longer tell the whole story.

A buyer can now ask ChatGPT, Gemini, Perplexity, Claude, Google AI Overviews, or another AI system which company to trust, which product to buy, which agency to hire, or which service to compare.

The answer may mention your brand.

It may recommend your competitor.

It may cite your website.

It may summarize your category without including you at all.

That is why brands need a new measurement layer.

That layer is Share of Model Voice, or SOMV.

Share of Model Voice measures how often and how strongly your brand appears inside AI-generated answers compared with competitors.

In plain English:

SOMV shows whether AI systems include your brand in the answers your buyers trust.

Traditional SEO asks:

Where do we rank?

SOMV asks:

When AI systems answer buyer questions, how often do we show up, how accurately are we described, and how strongly are we recommended?

That is the measurement shift inside SEO 3.0.

In the Blue Ninja system, Authority Infrastructure is the category. SEO 3.0 is the operating model. The Auth Graph is the strategy map. EntityMesh builds the infrastructure. EntityAgent answers from the approved knowledge base. SOMV measures model visibility. EchoScan monitors what AI systems and the web reflect back.

Table of Contents

• What is Share of Model Voice?
• Why does SOMV matter?
• How is SOMV different from traditional share of voice?
• How is SOMV different from SEO rankings?
• What should SOMV measure?
• What is the basic SOMV formula?
• Why should SOMV be weighted?
• Which prompts should you track?
• Which AI systems should you monitor?
• How often should you measure SOMV?
• What does a low SOMV score mean?
• How do you improve SOMV?
• How do EntityMesh, EntityAgent, SOMV, and EchoScan fit in?
• What should you do next?
• The bottom line
• Frequently asked questions
• Sources

What is Share of Model Voice?

Share of Model Voice, or SOMV, is a measurement framework for tracking how often, how prominently, and how positively a brand appears inside AI-generated answers for important prompts.

It is the AI search version of share of voice.

But instead of measuring visibility in ads, media, social conversations, or traditional search rankings, SOMV measures visibility inside model-generated answers.

SOMV can answer questions like:

• Does ChatGPT mention our brand for buyer-intent prompts?
• Does Gemini recommend us or our competitors?
• Does Perplexity cite our website?
• Does Google AI Overview use our pages as sources?
• Are we mentioned first or buried after competitors?
• Are we described accurately?
• Are we cited as a trusted source?
• Are we included in comparisons?
• Are competitors appearing more often than us?
• Are AI systems connecting us to the right category?
• Are we visible for high-intent prompts or only generic prompts?

SOMV matters because AI-generated answers are becoming part of the buyer journey.

A 2026 Search Engine Land guide defines share of AI citations as the portion of brand mentions in AI-generated answers that belongs to your brand compared with competitors. That is very close to the basic idea behind SOMV, but Blue Ninja expands the metric beyond mention count to include position, citation quality, sentiment, accuracy, and recommendation strength. (Search Engine Land)

A simple mention count is useful.

A weighted visibility model is better.

Want to see whether your brand is showing up in AI answers? Run the free EntityMesh scan.

Why does SOMV matter?

SOMV matters because AI systems can influence buying decisions before a visitor ever reaches your website.

A person might ask:

• "What are the best companies for AI search visibility?"
• "Which agency understands SEO 3.0?"
• "Who should I hire to improve visibility in ChatGPT?"
• "What are the best tools for tracking AI mentions?"
• "What is the best alternative to traditional SEO?"
• "Which product is better for my use case?"

If the answer recommends your competitor, your buyer's shortlist may form without you.

That is the new risk.

A brand can have good rankings and still lose influence inside AI answers.

A 2026 study of Google AI Overviews found that nearly 30% of AI Overview-cited domains did not appear in the co-displayed first-page organic results. It also found that AI Overviews appeared far more often for question-form queries than many other query types. That matters because many commercial AI prompts are phrased as questions. (arXiv)

Another 2026 study comparing Google Search, Gemini, and AI Overviews found low overlap between source retrieval patterns across different systems, which means visibility in one environment does not guarantee visibility in another. (arXiv)

SOMV matters because the buyer journey is fragmenting.

Traditional analytics may tell you who clicked.

SOMV helps show who was considered.

That distinction matters.

How is SOMV different from traditional share of voice?

Traditional share of voice usually measures how visible a brand is compared with competitors across channels like advertising, media, PR, social mentions, or search impressions.

SOMV measures something more specific:

How much of the AI-generated answer space does your brand own?

Traditional share of voice might ask:

• How often are we mentioned in the press?
• How much ad impression share do we have?
• How often are we discussed on social media?
• How much branded search demand do we capture?
• How many backlinks or citations do we have?

SOMV asks:

• How often do AI systems mention us?
• How often do they recommend us?
• Are we cited as a source?
• Are we described correctly?
• Do we appear before competitors?
• Do we appear in high-intent buyer prompts?
• Which sources are used to support the answer?
• Which competitors are present in the same answer?
• Does the AI explanation match our positioning?

That makes SOMV a more focused metric for AI Search Visibility.

Traditional share of voice measures market presence.

SOMV measures model-mediated presence.

In SEO 3.0, both matter.

But SOMV helps answer the question most brands are starting to ask:

When AI systems talk about my category, am I part of the answer?

How is SOMV different from SEO rankings?

SEO rankings measure where a page appears in a traditional search result.

SOMV measures whether a brand appears inside AI-generated answers.

Those are related, but not identical.

A page can rank without the brand being recommended.

A brand can be mentioned in an AI answer without being the top organic result.

A source can be cited by an AI system even if it is not in the first few traditional links.

This is why rankings are no longer enough.

Google says site owners should continue following Search fundamentals for visibility in AI features, including helpful content and technical accessibility. That confirms SEO still matters, but it also makes clear that AI features are now part of the broader search experience. (Google for Developers)

Here is the practical difference:

SEO tells you whether people can find your page.

SOMV tells you whether AI systems include your brand in the answer.

You need both.

What should SOMV measure?

SOMV should measure more than whether your brand is mentioned.

A brand mention can be weak or strong.

For example, these are not equal:

• Your brand is listed first as the best option.
• Your brand is mentioned third with no explanation.
• Your brand is cited as a source.
• Your brand is described incorrectly.
• Your competitor is recommended, and your brand is omitted.
• Your brand appears in an informational answer but not in buyer-intent recommendations.
• Your brand is mentioned, but only as one of many options.
• Your brand is cited for a definition, but not connected to the service you sell.

That is why SOMV should include several dimensions.

Mention frequency

How often does your brand appear across target prompts?

Mention position

Does your brand appear first, second, third, or only near the end?

Recommendation strength

Is the AI system actively recommending your brand, or merely listing it?

Citation presence

Is your website cited as a source?

Citation quality

Which page is cited, and is it the page you want AI systems to use?

Answer accuracy

Is your brand described correctly?

Sentiment

Is the mention positive, neutral, cautious, or negative?

Competitor proximity

Which competitors appear alongside your brand?

Prompt intent

Does the brand appear in high-intent buyer prompts or only broad educational prompts?

Source diversity

Are mentions supported by multiple credible sources or only by your own site?

A serious SOMV model should combine these signals instead of treating every mention as equal.

EntityMesh can help identify the missing assets behind weak SOMV, including unclear categories, missing proof, thin FAQs, and weak comparison infrastructure. Run the scan.

What is the basic SOMV formula?

The simplest SOMV formula is:

``text SOMV = Your brand mentions in AI answers / Total relevant brand mentions in those same answers ``

For example, imagine you track 100 high-intent prompts across ChatGPT, Gemini, Perplexity, and Google AI Overviews.

Across those answers, all brands in your competitor set are mentioned 200 total times.

Your brand is mentioned 30 times.

Your basic SOMV is:

``text 30 / 200 = 15% ``

That means your brand owns 15% of the model-visible brand mentions across that prompt set.

This basic formula is useful because it gives you a starting benchmark.

But it is incomplete.

It treats every mention as equal.

A first-place recommendation is worth more than a passing mention.

A cited recommendation is worth more than an uncited list item.

An accurate description is worth more than a vague or wrong one.

A high-intent prompt is worth more than a low-intent prompt.

So basic SOMV is a good starting point, but weighted SOMV is better.

Why should SOMV be weighted?

SOMV should be weighted because AI answer visibility is not binary.

You are not simply visible or invisible.

You can be:

• Strongly recommended
• Weakly mentioned
• Accurately described
• Incorrectly described
• Cited
• Uncited
• First in the list
• Last in the list
• Present in informational prompts
• Absent from buying prompts
• Mentioned with competitors
• Mentioned without differentiation

Those differences matter.

A better SOMV model should assign weight to signals like:

For example:

A brand mentioned first with a citation in a high-intent prompt should receive more weight than a brand mentioned once in a long generic list.

Weighted SOMV makes the metric more honest.

The goal is not to inflate the number.

The goal is to understand whether AI systems are actually helping buyers consider your brand.

Which prompts should you track?

SOMV is only as useful as the prompt set behind it.

If you track the wrong prompts, you get the wrong answer.

A good prompt set should include the questions your buyers actually ask.

Category prompts

These measure whether your brand is associated with the right category.

Examples:

• "Best AI search visibility companies"
• "Top SEO 3.0 agencies"
• "Who helps businesses show up in ChatGPT?"
• "Best companies for answer engine optimization"

Problem prompts

These measure whether your brand appears when buyers describe pain.

Examples:

• "Why is my brand not showing up in AI search?"
• "Why does ChatGPT recommend my competitors?"
• "How do I get my company cited in AI answers?"
• "How do I improve visibility in Google AI Overviews?"

Comparison prompts

These measure decision-stage visibility.

Examples:

• "Best alternative to traditional SEO agency"
• "AEO vs GEO vs SEO 3.0"
• "Compare AI search visibility tools"
• "EntityMesh vs traditional content marketing"

Brand prompts

These measure accuracy.

Examples:

• "What does [brand] do?"
• "Is [brand] good for [use case]?"
• "Who is [brand] best for?"
• "Compare [brand] and [competitor]."

Local or niche prompts

These matter when location, industry, or use case affects buying.

Examples:

• "Best AI search visibility agency for local service businesses"
• "AI search visibility for law firms"
• "SEO 3.0 strategy for healthcare companies"
• "AI visibility tools for B2B SaaS"

The prompt set should be reviewed regularly because buyer language changes.

AI search is dynamic.

Your measurement should be too.

Which AI systems should you monitor?

You should monitor the systems your buyers actually use.

For many brands, that includes:

• ChatGPT
• Google AI Overviews
• Google AI Mode where available
• Gemini
• Perplexity
• Claude
• YouTube search
• Reddit search
• TikTok search
• Traditional Google search

The exact mix depends on your market.

B2B buyers may use ChatGPT, Perplexity, LinkedIn, Google, and industry communities.

Local buyers may use Google, Maps, AI Overviews, reviews, Reddit, TikTok, and YouTube.

Software buyers may use ChatGPT, Gemini, Perplexity, G2, Reddit, YouTube, comparison pages, and review sites.

A 2026 Business Insider report on CMOs adapting to AI platforms highlighted the growing importance of representation across influential platforms such as Reddit and YouTube, not just owned websites. (Business Insider)

That is why SOMV should not be limited to one model.

A brand may perform well in one system and poorly in another.

The goal is not to chase every platform equally.

The goal is to measure the places that influence your buyers.

How often should you measure SOMV?

SOMV should be measured regularly because AI responses change.

At minimum, brands should measure monthly.

Fast-moving categories may need weekly tracking.

High-competition categories may need even more frequent monitoring.

Manual tracking can work at first.

A simple spreadsheet might include:

• Date
• AI system
• Prompt
• Brand mentioned?
• Competitors mentioned
• Mention position
• Citation URL
• Sentiment
• Accuracy
• Recommendation strength
• Notes
• Screenshot or saved answer

More advanced tracking can use automated workflows or monitoring platforms.

TechRadar's June 30, 2026 guide on AI search visibility tracking recommends ongoing monitoring because AI responses are dynamic and may change across tools, contexts, and time. It also notes that brands can start with manual prompt logging before graduating to automated workflows or paid platforms. (TechRadar)

The key is consistency.

If your prompt set changes every week, your trendline will be noisy.

Use a stable core prompt set, then add exploratory prompts as needed.

What does a low SOMV score mean?

A low SOMV score means AI systems are not including your brand often enough or strongly enough across the prompts that matter.

But the reason can vary.

A low SOMV score could mean:

Your category is unclear

AI systems do not know where to place you.

Your proof is weak

You make claims, but the public evidence is thin.

Your content is too generic

You answer broad topics but not buyer-specific questions.

Your competitors have stronger infrastructure

They may have better comparison pages, reviews, third-party mentions, or answer hubs.

Your website is hard to use as a source

Pages may be unclear, outdated, poorly structured, or difficult to crawl.

Your third-party footprint is thin

The broader web may not confirm your positioning.

Your brand is described inconsistently

Different sources may explain your business in conflicting ways.

Your prompt set is too ambitious

You may be measuring prompts where your brand does not yet have the authority to appear.

This is why SOMV should not be treated as a vanity score.

It should be treated as a diagnostic.

Low SOMV shows where the market and the model do not yet understand you. EchoScan can monitor those changes over time, and EntityMesh can turn the weak spots into build work.

How do you improve SOMV?

You improve SOMV by building better Authority Infrastructure.

That means improving the evidence system behind your brand.

1. Clarify your category

Use clear language that matches how buyers ask questions.

Do not hide the category behind vague positioning.

2. Build an Auth Graph

An Authority Infrastructure Graph, or Auth Graph, maps the entities, proof points, relationships, sources, comparisons, and crawlable assets that determine whether AI systems can understand and recommend your brand.

The Auth Graph shows what you need to be known for and what infrastructure is missing.

3. Create direct answer assets

Build pages that answer real buyer questions clearly.

Examples:

• Glossary pages
• FAQ hubs
• Service pages
• Comparison pages
• Case studies
• Source-backed guides
• Product pages
• Pricing or process pages where appropriate

4. Add proof

Turn claims into evidence.

Publish case studies, examples, screenshots, data, testimonials, and expert explanations.

5. Build comparison infrastructure

If buyers compare options, help them compare accurately.

Do not let competitors define the category alone.

6. Strengthen third-party validation

Earn and organize mentions across relevant sources.

That may include directories, partner pages, reviews, podcasts, guest articles, YouTube, local press, Reddit, and niche communities.

7. Improve technical and semantic structure

Make content easier to crawl, extract, cite, and understand.

Use clear headings, schema where appropriate, internal links, concise definitions, and accessible page structures.

8. Monitor and iterate

SOMV is not one-and-done.

Measure the prompts.

Find the gaps.

Build the assets with EntityMesh.

Answer from approved knowledge with EntityAgent.

Check again.

That is the SEO 3.0 measurement loop.

EntityMesh helps turn SOMV gaps into a build plan. Run the free scan to see where your Authority Infrastructure needs work.

How do EntityMesh, EntityAgent, SOMV, and EchoScan fit in?

SOMV is the measurement metric.

EchoScan is the monitoring layer.

EntityMesh is the infrastructure product.

EntityAgent is the approved-knowledge answer layer.

Here is the Blue Ninja system:

Authority Infrastructure

The service category. This is the structured, crawlable, source-backed knowledge system Blue Ninja builds.

SEO 3.0

The operating model. It explains how modern search now spans traditional SEO, AI answers, social search, answer engines, and future agents.

Auth Graph

The strategy framework. It maps what the brand needs to be known for, what proof supports it, and which gaps prevent stronger visibility.

EntityMesh

The infrastructure product. It turns the Auth Graph into live assets such as answer hubs, FAQ systems, service pages, glossary pages, comparison pages, internal links, schema-ready content, and approved brand answers.

EntityAgent

The answer layer. It retrieves from the approved, versioned EntityMesh knowledge base so public answers stay consistent and supported.

SOMV

The measurement metric. It tracks whether your brand appears, is cited, is accurately described, and is recommended inside AI-generated answers.

EchoScan

The monitoring layer. It tracks what search engines, AI systems, and the broader web reflect back about your brand over time.

The system works like this:

Map the Auth Graph.

Build with EntityMesh.

Answer from approved knowledge with EntityAgent.

Measure SOMV.

Monitor with EchoScan.

Improve the Authority Infrastructure.

That is how AI Search Visibility becomes measurable.

What should you do next?

Start simple.

You do not need a complicated dashboard on day one.

Step 1: Choose 20 high-value prompts

Pick prompts that represent your real buyer journey.

Include category, problem, comparison, brand, and use-case prompts.

Step 2: Choose your competitor set

Pick the brands you actually compete with for attention, trust, and recommendations.

Step 3: Run the prompts across multiple systems

Use ChatGPT, Gemini, Perplexity, Claude, Google AI Overviews where available, and traditional Google.

Step 4: Track mentions and recommendations

Log whether your brand appears, where it appears, whether it is cited, and how strongly it is recommended.

Step 5: Review answer accuracy

Do AI systems understand what you do?

Do they describe you correctly?

Do they recommend you for the right use cases?

Step 6: Identify missing infrastructure

Look for gaps in category clarity, proof, comparison content, FAQ coverage, third-party validation, internal linking, and source quality.

Step 7: Build from the gaps

Do not create random content.

Build the infrastructure that closes the SOMV gap.

That is how measurement becomes strategy.

The bottom line

SEO rankings still matter, but they no longer capture the full visibility picture.

AI systems are becoming part of the buyer journey.

They summarize markets.

They recommend competitors.

They cite sources.

They answer questions without always sending traffic.

That means brands need a way to measure whether they are part of the answer.

That metric is Share of Model Voice.

SOMV helps you understand how often and how strongly your brand appears inside AI-generated answers compared with competitors.

But SOMV is not just a score.

It is a diagnostic.

It shows where your Authority Infrastructure is strong, weak, or missing.

If you want to improve AI Search Visibility, start by measuring the answer space.

Then build the infrastructure that earns your place in it.

Run the free EntityMesh scan to start identifying the authority gaps behind your current Share of Model Voice.

Frequently asked questions

What is Share of Model Voice?

Share of Model Voice, or SOMV, measures how often and how strongly a brand appears inside AI-generated answers compared with competitors. It tracks brand mentions, citations, recommendation strength, answer accuracy, sentiment, position, and visibility across important prompts.

What does SOMV stand for?

SOMV stands for Share of Model Voice. It is Blue Ninja's measurement metric for tracking a brand's visibility inside AI-generated answers and AI-assisted discovery systems.

How is Share of Model Voice calculated?

A simple SOMV formula is your brand mentions in AI answers divided by total relevant competitor mentions in those same answers. More advanced SOMV models weight mentions by position, citation quality, sentiment, accuracy, recommendation strength, prompt intent, and commercial value.

Why is SOMV important?

SOMV is important because buyers increasingly use AI systems to ask for recommendations, comparisons, and trusted providers. Traditional SEO metrics show rankings and clicks, but SOMV shows whether a brand is included, cited, and recommended inside AI-generated answers.

How is SOMV different from SEO rankings?

SEO rankings measure where a page appears in traditional search results. SOMV measures how often and how strongly a brand appears inside AI-generated answers. A page can rank without the brand being recommended, and a brand can be mentioned by AI without holding the top organic position.

What prompts should I track for SOMV?

Track prompts that reflect the buyer journey, including category prompts, problem prompts, comparison prompts, brand prompts, and use-case prompts. Examples include "best companies for [category]," "who helps with [problem]," "compare [brand] and [competitor]," and "is [brand] good for [use case]?"

Which AI systems should I monitor for SOMV?

Monitor the AI and discovery systems your buyers use. For many brands, that includes ChatGPT, Google AI Overviews, Google AI Mode where available, Gemini, Perplexity, Claude, YouTube, Reddit, TikTok, and traditional Google search.

What is a good SOMV score?

A good SOMV score depends on your category, competitor set, prompt intent, and current authority. Early-stage brands may start with low SOMV, while established brands should expect stronger visibility across branded, category, and comparison prompts. The goal is to improve SOMV over time, especially for high-intent prompts.

How do you improve Share of Model Voice?

You improve SOMV by building stronger Authority Infrastructure. That includes clearer category language, better answer assets, stronger proof, comparison pages, third-party validation, structured content, internal linking, approved answer infrastructure through EntityAgent, and ongoing monitoring through EchoScan.

How does EchoScan relate to SOMV?

EchoScan is Blue Ninja's monitoring layer for tracking what search engines, AI systems, and the broader web reflect back about a brand. SOMV is one of the key metrics EchoScan can monitor because it measures how often and how strongly the brand appears inside AI-generated answers.

How does EntityAgent relate to SOMV?

EntityAgent can reduce answer inconsistency by serving responses from the approved, versioned EntityMesh knowledge base. It does not replace SOMV measurement, but it helps create a cleaner owned answer surface for buyers, customers, crawlers, and AI agents.

Sources

• Search Engine Land: How to measure brand visibility in AI search
• Google Search Central: A new resource for optimizing for generative AI in Google Search
• Measuring Google AI Overviews, activation, cited domains, source quality, and publisher impact
• How Generative AI Disrupts Search: Google Search, Gemini, and AI Overviews
• TechRadar: How to track your brand's visibility in AI search results
• Business Insider: How CMOs are adapting brand strategy for AI platforms