How Share of Voice is calculated
In each AI response, multiple domains get cited as sources. Share of Voice measures how many of those citations belong to you:- yoursite.com ← yours
- g2.com
- techcrunch.com
- competitor.com
- wikipedia.org
How we report Share of Voice
| Level | What it shows |
|---|---|
| Per response | Your share in a single AI answer |
| Per prompt | Average share across all responses for that prompt |
| Dashboard (overall) | Average share across all your tracked prompts |
| Per page | How often a specific page from your domain gets cited |
What counts as a “competitor” in AI search
In traditional SEO, competitors are other businesses in your market. In AI search, competitors are any domain that AI cites instead of yours. If you’re Pipedrive (a CRM), domains competing for your share of voice include:- Direct competitors — hubspot.com, salesforce.com, zoho.com
- Review sites — g2.com, capterra.com (if AI cites them instead of you)
- Wikipedia — wikipedia.org (if AI pulls generic info instead of citing your site)
- Comparison blogs — techradar.com, pcmag.com
Share of Voice is calculated based on citations, not brand mentions. You can be mentioned by name but still lose SoV if AI cites competitor domains as sources.
Why you can’t always achieve 100% SoV
For high-value, commercial searches, AI models are designed to provide balanced, helpful answers — not to promote a single brand. This means:- AI cites third-party sources for credibility — Review sites, comparison articles, and industry publications add objectivity
- Multiple options help users — AI aims to be helpful, which often means presenting choices
- Bias detection — AI models are trained to avoid appearing to endorse any single brand too heavily
| Market Position | Realistic SoV Target |
|---|---|
| Market leader | 35-50% |
| Strong contender | 20-35% |
| Niche player | 15-25% (in your niche) |
| New entrant | 5-15% |
What sources each AI model prefers
Different AI platforms have different preferences for where they pull information. Understanding this helps you prioritize your efforts:| Platform | Preferred Sources | How to Improve SoV | Update Frequency |
|---|---|---|---|
 ChatGPT | Wikipedia, major publications (TechCrunch, Forbes), Reddit, review sites (G2, Capterra) | Get Wikipedia presence, coverage in major publications, build Reddit presence | Training data + browse |
 Claude | In-depth documentation, research publications, detailed guides, professional journals | Create well-reasoned content, publish detailed guides, focus on educational over promotional | Training data focused |
 Perplexity | Fresh content, news articles, comparison articles, structured pages | Keep content updated, use clear headings, publish timely industry content | Real-time search |
 Gemini | SEO-optimized pages, Google Business profiles, YouTube, schema markup | Optimize for traditional SEO, use schema markup, create YouTube content | Google index |
 Grok | X/Twitter posts, real-time news, trending topics | Maintain active X presence, engage in industry conversations | Real-time |
 DeepSeek | Technical documentation, code repositories, developer content | Create technical content, maintain GitHub presence | Training data |
 Qwen | Chinese and international sources, e-commerce platforms, Alibaba ecosystem | Optimize for Chinese markets if relevant, e-commerce presence | Training data |
What drives Share of Voice
Brand recognition
Brand recognition
Well-known brands with consistent web presence tend to be mentioned more. AI aggregates information from across the web — the more you’re mentioned everywhere, the more AI mentions you.
Content coverage
Content coverage
Brands with more content covering more topics get mentioned in more conversations. Content breadth expands your SoV. If you don’t have a page about a topic, you won’t be mentioned for queries about it.
Third-party mentions
Third-party mentions
When industry publications, review sites, and comparison articles mention you positively, AI picks up on these references. Third-party validation is often weighted more heavily than your own content.
Relevance matching
Relevance matching
Your SoV is higher for queries that closely match your positioning and lower for queries outside your sweet spot. A CRM focused on agencies will have higher SoV for “best CRM for agencies” than “best enterprise CRM.”
Recency
Recency
Fresh content and recent mentions can boost SoV, especially on platforms like Perplexity that prioritize recency. Stale content leads to declining share.
Improving Share of Voice
Growing your SoV
Identify low-SoV categories
Filter prompts by category and find where your SoV is below your target. These are your growth opportunities.
Analyze what competitors do
For categories where competitors lead, understand what content and positioning they have. Check their cited sources.
Create targeted content
Build content specifically designed to capture share in weak categories. Match the depth and format of winning competitors.
Pursue third-party coverage
Get mentioned by publications and review sites that AI references. This often has more impact than your own content.
Defending your SoV
If your SoV is declining:- Check competitor activity — Did they launch new content or get new coverage?
- Verify content freshness — Is your content still current and accurate?
- Look for new sources — Are new publications or sites citing competitors?
- Review query shifts — Have the queries in your market changed?
SoV benchmarks
- Market leader
- Strong contender
- Emerging player
- Low visibility
SoV: 35-50%You dominate AI conversations in your category. Focus on:
- Defending position against challengers
- Expanding into adjacent categories
- Maintaining content freshness
- Monitoring for emerging competitors
Common SoV mistakes
Tracking too few competitors
Tracking too few competitors
If you only track 2-3 direct competitors, your SoV will look artificially high. Add review sites, comparison blogs, and indirect competitors for accurate measurement.
Ignoring third-party sources
Ignoring third-party sources
Your own content alone won’t maximize SoV. AI models trust third-party validation. Invest in getting coverage on sites AI cites.
Focusing only on branded queries
Focusing only on branded queries
Tracking “What is [Your Brand]?” will give you 100% SoV but tells you nothing useful. Focus on category queries where users haven’t decided yet.
Expecting instant results
Expecting instant results
SoV changes slowly. AI models update at different rates — some use training data, others search live. Give improvements 4-8 weeks to show impact.