How to Measure AI Share of Voice
Share of voice has been a core marketing metric for decades. In traditional media, it measures what percentage of the total advertising in a category belongs to your brand. In SEO, it measures what percentage of organic search clicks go to your website for a set of target keywords. Now, in the age of AI-powered search, share of voice takes on a new — and arguably more important — meaning.
AI share of voice measures how often AI platforms recommend or mention your brand relative to competitors when users ask questions in your category. It is the single most important metric for understanding your competitive position in AI-powered discovery.
This guide explains what AI share of voice is, how to calculate it, how to benchmark against competitors, and how to use it to drive strategic decisions.
What AI Share of Voice Means
When a user asks ChatGPT, Claude, Perplexity, or any other AI assistant a question like "What is the best project management tool for small teams?", the AI generates a response that typically mentions several brands. AI share of voice measures your brand's proportion of those mentions across a defined set of queries and platforms.
Unlike traditional share of voice, which is typically measured in a single channel (advertising spend, organic search traffic), AI share of voice spans multiple platforms. A user might ask the same question to ChatGPT, Perplexity, and Google Gemini and receive different brand recommendations each time. Your AI share of voice encompasses your presence across all of these platforms.
The concept is straightforward:
AI Share of Voice = (Number of times your brand is mentioned / Total number of brand mentions across all tracked brands) x 100
If you track 50 queries across five AI platforms and record a total of 200 brand mentions across all responses, and your brand accounts for 40 of those mentions, your AI share of voice is 20%.
Why AI Share of Voice Matters More Than You Think
AI share of voice is not merely a vanity metric. It has direct commercial implications:
It predicts discovery outcomes
When a potential customer asks an AI assistant for a recommendation, the brands mentioned in the response form their consideration set. If your competitor has a 35% share of voice and you have a 10% share, they are being recommended to roughly three and a half times as many AI users as you are. That gap compounds over time as AI adoption grows.
It captures a channel traditional metrics miss
Your Google Analytics will not show you the customer who asked ChatGPT for a recommendation, received your competitor's name, and went directly to their website. That customer never appeared in your organic search data. AI share of voice is the metric that captures this invisible competitive dynamic.
It reveals platform-specific strengths and weaknesses
Your overall share of voice might be 25%, but that average could mask significant variation: perhaps you have 40% share on Perplexity (because your content ranks well in web search) but only 10% on ChatGPT (because the model's training data favors a competitor). Platform-level share of voice data reveals exactly where to focus your efforts.
It provides a competitive benchmark
Absolute mention rates tell you how visible you are. Share of voice tells you how visible you are relative to the competition. This relative measure is what ultimately determines who wins the customer.
How to Calculate AI Share of Voice
Step 1: Define your query set
Select 20-50 queries that represent the questions your target customers ask when they are in the market for your product or service. These should include:
- Category queries: "What is the best [your category]?"
- Comparison queries: "[Your brand] vs [competitor]"
- Recommendation queries: "Can you recommend a [product/service type] for [use case]?"
- Problem-solving queries: "How do I [solve problem your product addresses]?"
- Feature-specific queries: "Which [category] has the best [key feature]?"
The quality of your query set determines the quality of your share of voice measurement. Choose queries that genuinely reflect how your customers search, not queries that you think will make your metrics look good.
Step 2: Select your platforms
At minimum, track across these five platforms:
| Platform | Why it matters |
|---|---|
| ChatGPT | Largest user base; highest volume of recommendation queries |
| Claude | Growing professional user base; strong in B2B contexts |
| Perplexity | Search-first platform with real-time web results; shows citations |
| Google Gemini | Integrated across Google ecosystem; reaches Android users |
| Google AI Overviews | Appears in Google search results; broadest reach |
Step 3: Run queries and record mentions
For each query on each platform, record:
- Which brands were mentioned in the response
- The order in which they were mentioned (first mention typically carries more weight)
- Whether the mention was a recommendation, a neutral reference, or a negative comment
Step 4: Calculate share of voice
Aggregate your data and calculate share of voice at three levels:
Overall share of voice:
Total mentions of your brand across all queries and platforms / Total mentions of all brands across all queries and platforms x 100
Platform-specific share of voice:
Your brand mentions on Platform X / All brand mentions on Platform X x 100
Query-category share of voice:
Your brand mentions for query category Y / All brand mentions for query category Y x 100
Platform-specific and category-specific breakdowns are often more actionable than the overall figure because they tell you exactly where you are winning and where you are losing.
Step 5: Weight by platform importance
Not all platforms carry equal weight for your business. If your target audience primarily uses ChatGPT, a 30% share of voice on ChatGPT matters more than a 50% share on a platform your audience rarely uses.
Consider creating a weighted share of voice that reflects platform importance:
Weighted SOV = Sum of (Platform SOV x Platform Weight) for each platform
Assign weights based on your audience's platform usage patterns. If you do not have precise data, a reasonable starting point for B2B businesses is:
| Platform | Suggested weight |
|---|---|
| ChatGPT | 35% |
| Google AI Overviews | 25% |
| Perplexity | 20% |
| Claude | 15% |
| Gemini | 5% |
Adjust these based on your industry and audience. Consumer-facing businesses might weight Google AI Overviews higher, while technology companies might weight Perplexity and Claude higher.
Benchmarking Against Competitors
Raw share of voice numbers are most useful when compared against competitors. Here is how to structure your competitive benchmarking:
Identify your benchmark set
Select 3-5 direct competitors to track alongside your brand. Include:
- Your most direct competitor (similar product, similar market)
- The market leader in your category (even if they are much larger)
- An emerging competitor who is growing quickly
- Optionally, an adjacent competitor from a related category
Create a competitive dashboard
Track share of voice over time for each competitor. A monthly cadence is sufficient for most businesses. Your dashboard should show:
- Overall SOV trend: A line chart showing each competitor's share of voice month over month
- Platform breakdown: A stacked bar chart showing share of voice by platform for each competitor
- Movement alerts: Flags when any competitor's share increases or decreases by more than 5 percentage points
Interpret the competitive data
Look for these patterns:
- A competitor's SOV is rising while yours is flat. This is a warning sign — they are investing in AI visibility and gaining ground. Investigate what they are doing differently.
- Your SOV is high on Perplexity but low on ChatGPT. This suggests your web content is strong (Perplexity uses real-time search) but your broader brand authority signals are weak (ChatGPT relies more on training data and overall brand reputation).
- A new competitor is appearing that was not in previous audits. This signals a new entrant who is aggressively building AI visibility. Monitor them closely.
- Your SOV is declining despite your efforts. This may indicate that the competitive landscape is intensifying and you need to accelerate your efforts, or that a specific optimization approach is not working.
Advanced Share of Voice Metrics
Beyond basic mention counting, consider these more nuanced metrics:
Weighted mention position
A first-position mention (your brand is named first in the AI's response) is more valuable than a third-position mention. Weight your share of voice by mention position:
- First mention: 3 points
- Second mention: 2 points
- Third or later mention: 1 point
Position-Weighted SOV = Your weighted points / Total weighted points across all brands x 100
Sentiment-adjusted share of voice
Not all mentions are equal. A positive recommendation is more valuable than a neutral mention, and a negative mention can be worse than no mention at all. Adjust your share of voice by sentiment:
- Positive mention: 1.5 points
- Neutral mention: 1 point
- Negative mention: -0.5 points
This prevents you from celebrating a high share of voice that is actually driven by negative commentary.
Query-type share of voice
Break your share of voice down by query intent:
- Recommendation queries: "What is the best..." — the highest commercial value
- Comparison queries: "[Brand] vs [Brand]" — high purchase intent
- Informational queries: "What does [brand] do?" — awareness-stage value
- Problem-solving queries: "How do I..." — solution-stage value
Tracking share of voice by query type reveals where you are strongest in the buyer's journey and where you are losing potential customers.
Tools and Methods for Tracking
Manual tracking
For initial audits or very small businesses, manual tracking works but has significant limitations:
- Time-consuming: 50 queries across 5 platforms = 250 individual checks
- Inconsistent: AI responses vary between sessions
- Difficult to maintain: Monthly re-runs require substantial effort
- No trend data: Hard to track changes over time without structured recording
Automated tracking
RivalScope automates the entire share of voice measurement process. It systematically queries all five major AI platforms, records every brand mention, calculates share of voice at the overall, platform, and query level, and tracks trends over time. This transforms share of voice from a periodic manual exercise into a continuously updated competitive intelligence metric.
Automated tracking is essential if you want to:
- Measure share of voice consistently over time
- Track more than 20-30 queries
- Monitor multiple competitors simultaneously
- Identify trends and respond to changes quickly
- Report share of voice to stakeholders with confidence
Using Share of Voice to Drive Strategy
Share of voice data should directly inform your AI visibility strategy:
- Low overall SOV: Focus on foundational activities — building authoritative content, earning third-party mentions, and improving brand entity signals across the web.
- Low SOV on specific platforms: Investigate why. If ChatGPT is a gap, focus on earning mentions in sources that influence its training data. If Perplexity is a gap, focus on SEO and structured content that ranks well in web search.
- Declining SOV despite steady effort: Re-evaluate your approach. Your competitors may be outpacing you, or the queries your audience uses may be shifting.
- High SOV but low conversion: Your AI visibility is strong, but the mentions may not be driving action. Focus on improving how AI platforms describe your brand — ensuring they mention your key differentiators and direct users toward conversion.
Share of voice is not a destination metric. It is a diagnostic metric that tells you where to invest next and whether your investments are paying off.
Measure your AI share of voice across every platform -- start a free 3-day trial with RivalScope.