Share of voice calculation: how PR teams should measure visibility without fooling themselves
Learn how to calculate share of voice across PR, social, search, and AI answers, with formulas, reporting tips, and common measurement mistakes.
Share of voice calculation is one of the simplest ways to answer a hard question: are we showing up more than the competitors our buyers already know?
For PR and B2B marketing teams, that question now stretches beyond press mentions. It includes search visibility, social conversation, analyst and newsletter mentions, and increasingly, whether AI answer engines cite or describe your brand at all.
Table of contents
Jump to each section:
- What share of voice calculation actually measures
- The basic share of voice formula
- How to calculate share of voice by channel
- What to include in a useful SOV report
- Common mistakes that make SOV misleading
- How to improve share of voice without inflating the metric
- How share of voice changes in AI search
- A practical workflow for PR and marketing teams
What share of voice calculation actually measures
Share of voice, often shortened to SOV, measures your brand's visibility compared with a defined competitor set. In PR, that usually means your share of relevant media mentions. In social, it may mean your share of brand or category conversations. In search, it can mean your estimated share of organic visibility for a keyword set.
The useful part is not the percentage by itself. The useful part is the comparison. A startup with 9% SOV in a niche category may be punching above its weight if the rest of the conversation is split among large incumbents. A market leader with 9% SOV may be losing narrative control if two challengers are consistently cited more often.
This is why the denominator matters. You are not measuring all mentions on the internet. You are measuring your share of a defined market conversation during a defined period, across sources that matter to the audience you want to influence.
Ariana Sorensen
The basic share of voice formula
The simplest formula is:
Share of voice = Brand mentions / Total category mentions x 100
If your brand earned 120 relevant mentions and your competitor set earned 880 relevant mentions in the same period, the total category mention count is 1,000. Your SOV is 12%.
That formula is consistent with how major monitoring vendors explain the metric. Cision frames SOV as brand mentions divided by total industry mentions, multiplied by 100. Mention uses the same basic structure, comparing your mentions with your brand plus competitor mentions.
The formula gets more useful when you define three inputs before collecting data:
- Competitor set: Pick the three to eight brands your buyer would realistically compare you with.
- Source set: Decide whether the calculation includes news, trade publications, blogs, social, podcasts, search, AI answers, or paid media.
- Time window: Use a window that matches the decision you want to make, such as launch week, campaign month, quarterly planning, or annual benchmark.
Without those inputs, SOV becomes a decorative number. With them, it becomes a planning tool.
How to calculate share of voice by channel
A single SOV number can hide more than it reveals. PR teams should calculate SOV by channel first, then roll it up only when the data quality is comparable.
For earned media SOV, count relevant editorial mentions across the publications that influence your audience. Relevance should be stricter than keyword matching. A passing mention in a syndicated list is not equal to a feature story that explains your positioning, quotes your executive, and links to your research.
For social SOV, count brand mentions, tagged mentions, relevant untagged mentions, and category conversations where your brand appears beside competitors. Sprout Social notes that SOV can include social, media, paid, and search contexts, but also warns that qualitative factors such as relevance and impact matter.
For search SOV, use a fixed keyword set and compare the estimated visibility or click share your domain receives against competitors. This works best for evergreen topics where rankings are relatively stable. For volatile news queries, search SOV can swing too quickly to support strategy decisions.
For paid SOV, use impression share from ad platforms. This is the cleanest version mathematically because the ad platform can tell you how often your ad showed compared with eligible impressions. It is also the least useful for PR unless you are comparing earned visibility against paid support during the same campaign.
The best PR reports keep these channels separate, then explain the relationship. A product launch might have low media SOV but high LinkedIn SOV among a target buyer segment. A data report might have modest social conversation but strong trade media visibility. Both outcomes can be useful, but they point to different next steps.
What to include in a useful SOV report
A good SOV report should make it easy for a leadership team to understand what changed, why it changed, and what to do next. The percentage is only the entry point.
Include the following:
- Overall SOV: Your brand's share against the chosen competitor set.
- SOV by channel: Earned media, social, search, AI answers, paid, or any channel you can measure consistently.
- SOV by topic: Which narratives, product lines, issues, or audience needs are driving visibility.
- Sentiment and message pull-through: Whether the visibility is positive, neutral, negative, or aligned with the story you intended to tell.
- Source quality: Which publications, creators, newsletters, communities, or answer engines are driving the number.
- Change over time: A comparison against the previous period, campaign baseline, or competitor benchmark.
This is where SOV becomes more valuable than raw coverage volume. Fifty mentions may look strong until competitors earned 900 in the same category. A 20% SOV may look healthy until sentiment shows that half the attention came from a support issue or executive controversy.
ContentGrip Staff
Common mistakes that make SOV misleading
The first mistake is counting every mention equally. A homepage feature in a trusted trade publication is not the same as a scraped quote on a low-quality aggregator. Weighting is imperfect, but ignoring source quality is worse.
The second mistake is using an unstable competitor set. If you compare against three competitors in January and eight in March, your SOV will change even if your actual visibility does not. Keep the core competitor set stable, then create separate views for emerging challengers.
The third mistake is blending unrelated channels into one score. Social conversation, organic search visibility, paid impression share, and AI citations are all visibility signals, but they are produced by different systems. A combined score can be useful for executives, but only after the underlying channel numbers are clear.
The fourth mistake is treating SOV as a proxy for trust. Visibility is not the same as credibility. A brand can dominate conversation for the wrong reason. That is why sentiment, message pull-through, audience fit, and source authority should sit beside SOV in the same report.
Tommy Prayoga, Head of Agency at digital PR service provider Content Collision: "Share of voice is useful only when the team can explain what kind of voice it is. A spike from controversy, a spike from low-quality syndication, and a spike from credible industry coverage should never be celebrated the same way. The PR job is not to win the loudest chart. It is to earn the right conversation in front of the right audience."
How to improve share of voice without inflating the metric
Improving SOV starts with deciding which conversation you want to win. Most mid-market B2B brands cannot dominate every broad category conversation. They can, however, own narrower topics that matter to a specific buying committee.
A useful approach is to build a topic map around the problems your buyer already researches. For example, a cybersecurity company may struggle to win general "cybersecurity" conversation, but it may build a stronger share around "software supply chain risk" or "AI phishing readiness." The narrower topic gives PR, content, analyst relations, and executive thought leadership a shared target.
Then, create assets that deserve citation. Proprietary data, expert commentary, customer evidence, technical explainers, and clear definitions all give journalists, newsletters, analysts, creators, and AI systems something concrete to reference. Generic brand claims rarely create durable SOV because they do not add new information to the market.
Finally, treat distribution as a system. A strong media pitch can create earned coverage. A clear executive LinkedIn post can extend the conversation. A structured evergreen article can capture search demand. A partner webinar can reinforce authority with a niche audience. The point is not to repeat the same message everywhere. The point is to make the same strategic narrative discoverable across the places your audience looks for proof.
How share of voice changes in AI search
AI search makes SOV harder and more important. In traditional search, visibility was usually measured by rankings, impressions, clicks, and traffic. In AI answers, visibility may mean whether your brand is mentioned, cited, summarized accurately, or excluded from the answer entirely.
Muck Rack's May 2026 research on what AI is reading found that earned media accounts for 84% of all AI citations, while paid and advertorial content accounts for 0.3%. That does not mean every brand should chase press for the sake of press. It means third-party validation is becoming part of the discovery infrastructure.
For PR teams, the AI version of SOV should track at least four signals:
- Mention rate: How often your brand appears in answers to relevant prompts.
- Citation rate: How often your owned or earned sources are cited.
- Sentiment and framing: Whether the answer describes your brand accurately and favorably.
- Competitor presence: Which competitors appear when you do not.
Patrecia Meliana
A practical workflow for PR and marketing teams
Start with a quarterly benchmark. Pick one category conversation, five competitors, three channel types, and a 90-day window. Calculate SOV for that baseline before launching a major campaign, not after.
Next, tag the data by topic. Do not only ask whether your brand was mentioned. Ask which narrative drove the mention. Was it product innovation, funding, customer proof, pricing, leadership, trust, or a category trend? This shows whether your visibility supports the positioning you actually want.
Then, review quality before volume. Separate tier-one editorial mentions, niche trade coverage, newsletters, social posts, analyst commentary, owned content citations, and AI answer appearances. A smaller number of high-quality mentions can be more strategically useful than a larger number of shallow mentions.
After each campaign, compare the new SOV with the benchmark. Look for three movements:
- Did total SOV increase against the same competitor set?
- Did the right topics drive the increase?
- Did the increase appear in the channels that influence your audience?
If the answer is yes across all three, the campaign likely improved meaningful visibility. If only the first answer is yes, you may have created noise rather than advantage.
The strongest SOV programs are boring in the best way. They use the same competitor set, the same measurement windows, the same source rules, and the same quality checks over time. That consistency is what makes the metric useful.
Share of voice calculation should not be treated as a vanity score. Used well, it is a disciplined way to see whether your PR, content, social, and AI visibility work is actually changing your position in the market conversation. Used lazily, it is just another percentage in a slide deck.
