Share of voice is one of the most familiar metrics in communications, and one that is easy to misuse. Teams report it without defining their competitive set. They average it across topics moving in opposite directions. They present it to executives as evidence of performance when it is really a measure of volume.
This guide covers what share of voice in PR actually measures, how to calculate it correctly, what benchmarks look like in practice, and why SoV belongs inside a broader analytics program rather than standing alone as a verdict.
What does share of voice in PR actually measure?
Share of voice is the proportion of qualifying media coverage your brand generates relative to a defined competitive set. It answers one question: of all the coverage being written about companies in your market, how much of it is about you? It measures volume, not quality. A high share of voice built on neutral or negative coverage is not a competitive advantage, and SoV is always relative to the competitive set behind it. A number without that context means nothing.
How do you calculate share of voice correctly?
Your brand's qualifying coverage divided by the total brand-level qualifying coverage across the competitive set, times 100. The formula is simple. The methodology is not. The most consequential choices are how you define the competitive set, what counts as qualifying coverage, whether to weight by outlet quality or audience, and whether to segment by topic. Changing the competitive set midstream makes period-over-period comparisons meaningless.
What does a good share of voice number look like?
There is no universal benchmark. Share of voice is always relative to the category, the size of the competitive set, and the topic being measured. Trajectory often matters more than the level: a brand moving from 8% to 14% over two quarters in a competitive market is a stronger signal than one holding flat at 28%. Topic-level SoV is usually more actionable than the aggregate, since a brand can hold strong overall SoV while sitting weak on the topics most important to its strategic position.
Why is share of voice not enough on its own?
Volume is not sentiment. A brand can dominate SoV while its coverage runs mostly negative or focused on non-strategic topics. SoV does not measure narrative: two brands can hold equal SoV while one owns the innovation story and the other owns the controversy story. SoV also varies sharply by channel in ways that aggregating hides. And traditional SoV does not capture AI visibility, where your brand's presence in AI-generated answers is becoming an increasingly consequential measure of competitive position.
What questions should comms teams ask any vendor about share of voice tracking?
What is in the denominator and does the source coverage match where your relevant coverage actually appears? Can you define your own competitive set or is it fixed by the platform? Can you segment by topic, channel, or time period? How does it weight coverage, by outlet quality, audience, or raw counts? And does it track AI answer engine share of voice, measuring how often your brand appears in AI-generated responses against competitors?
A team presents to the executive committee: "We hold 34% share of voice in our category." Heads nod. Then someone asks the only question that matters: is that good?
Nobody in the room can answer, because share of voice on its own is not an answer. It is a starting point that needs a competitive set, a topic breakdown, and a sentiment read before it means anything at all.
Share of voice, or SoV, is the percentage of total qualifying media coverage your brand generates relative to a defined competitive set.
It answers one question: of all the coverage being written about companies in your market, how much of it is about you?
Two qualifications belong in any working definition.
SoV measures volume, not quality. A high share of voice built on neutral or negative coverage is not a competitive advantage. A brand can dominate its category's coverage while its sentiment declines, its key messages get lost, and favorable coverage goes to competitors. Volume and quality are separate measurements, and they have to be read together.
SoV is always relative to a defined competitive set. The same brand might show 40% SoV in a five-competitor set and 8% in a twenty-competitor set. Both are mathematically correct. Neither is comparable to the other. A common error in SoV reporting is presenting a number without disclosing the competitive set behind it.
SoV = (your brand's qualifying coverage / total brand-level qualifying coverage across the competitive set) x 100
Qualifying coverage means the articles, clips, or placements in which each brand appears, counted consistently across the competitive set. The denominator is the sum of those brand-level coverage counts. A single piece can therefore count toward more than one brand if it mentions multiple competitors. That is different from counting every individual brand reference inside an article, which is share of mentions, a related metric covered below.
Worked example: your brand appears in 1,200 qualifying pieces of coverage over the quarter. Your four primary competitors appear in 900, 750, 600, and 550. Total brand-level qualifying coverage across the set: 4,000.
SoV = 1,200 / 4,000 x 100 = 30%.
Three complications decide whether that number is strategically useful.
1. Defining the competitive set. Which brands count as competitors is a strategic decision with methodological consequences. A narrow set of three or four direct rivals produces a higher SoV than a broad set covering every adjacent player. Neither is wrong, but they answer different questions. Define the set based on the decision the metric is meant to inform, then use that same set consistently across every reporting period. Changing it midstream makes period-over-period comparisons meaningless.
2. Weighting by outlet quality or audience. Raw counts treat a two-line mention in a small trade newsletter the same as a prominent feature in a publication that matters deeply to your audience. Some teams weight SoV by readership, outlet tier, or audience relevance. Whatever method you choose, apply it consistently and disclose the weighting.
3. Topic-level segmentation. Overall SoV often hides wide variation across topics. You might hold 30% overall while sitting at 5% on pricing conversations and 55% on product innovation. Those describe completely different competitive positions. Topic-level SoV is usually more actionable than the aggregate.
These terms are often used interchangeably across the industry. In this guide, we separate them because they answer different questions.
Share of voice is your proportion of brand-level qualifying coverage across the competitive set: how many qualifying pieces include your brand relative to the summed coverage counts for all brands. Share of mentions is how often your brand appears within coverage, including inside articles that name several brands. You can appear in 60% of the articles about your category while earning only 33% of the total brand mentions inside them, because competitors get referenced more often each time you show up together.
The distinction matters for narrative control. Article presence tells you where you show up. Mention frequency tells you whether you dominate the conversation once you are there. A brand with high SoV but low share of mentions is present in the category conversation without controlling it.
Delve measures both share of voice and share of mentions, alongside sentiment, for a brand and its competitors on one consistent data set, which is what lets you see the gap between showing up and dominating. For the research across 57 companies, including the four categories of narrative control and what separates Narrative Leaders from Narrative Underdogs, see Share of Voice vs. Share of Mentions: The Complete Guide.
There is no universal benchmark. Every category, competitive landscape, and business context produces a different SoV distribution. The ranges below are illustrative, not industry benchmarks. They show how the size of the competitive set changes the way you read a SoV number, and they should not be used as universal performance targets. Calibrate to your own set.
| Market position | Illustrative SoV range | Notes |
|---|---|---|
| Category leader 2 to 3 competitors | 35 to 55% | Illustrates what clear leadership can look like in a concentrated set |
| Category leader 5 to 10 competitors | 22 to 38% | Illustrates how leadership thresholds compress as the set expands |
| Strong challenger | 15 to 25% | Competitive with the leader and meaningfully present in the category conversation |
| Emerging brand | 8 to 15% | Visible in the conversation but not yet shaping it |
| New market entrant | 2 to 8% | Building from near zero, where growth rate matters more than the current level |
Three caveats on using these ranges.
Trajectory can matter more than level. A brand moving from 9% to 16% over three quarters is likely in a better position than one holding at 28% with flat or declining sentiment. Direction and rate of change often matter more than the current number.
Topic SoV is often more actionable than aggregate SoV. A brand can hold strong overall SoV while sitting weak on the two or three topics most important to its strategic position. Setting targets at the topic level produces more actionable measurement than the aggregate alone.
Volume without sentiment is incomplete. A brand gaining SoV through controversy or product issues is not in a better position than one with stable, lower SoV on favorable coverage. Always read SoV alongside sentiment trajectory. For the framework that puts SoV in context, see What Is PR Analytics? Metrics That Actually Measure Impact.
1. Volume is not sentiment. The most important limitation. A brand can dominate SoV while its coverage runs mostly negative, neutral, or focused on non-strategic topics. SoV tells you how loud you are. It says nothing about whether what is being said helps your position.
2. SoV does not measure narrative. Two brands can hold equal SoV while one owns the innovation story and the other owns the controversy story. SoV measures coverage volume, not what that coverage is doing to strategic position.
3. SoV varies by channel in ways that matter. Your SoV in tier-1 business press, trade publications, social, and analyst coverage can differ sharply. Aggregating across channels averages those differences away. Where specific audiences matter more than total volume, channel-segmented SoV is more useful than the aggregate.
4. Traditional SoV does not capture AI visibility. When your audiences research your brand, your competitors, or your category through AI tools, the SoV equivalent is how often your brand appears in AI-generated answers. That is measurable, and many traditional monitoring tools do not track it.
The definition of share of voice now has to extend to the AI-mediated information environment.
When a journalist asks ChatGPT about the leading platforms in a category, when a buyer asks Perplexity which vendors to evaluate, when an analyst asks Google's AI Mode about the competitive landscape, the answer becomes another surface where brand perception is formed. Your brand appearing in 60% of those responses versus 10% is a measurable and consequential difference.
AI share of voice differs from traditional SoV in three ways.
The denominator is a defined query set, not articles. You are measuring how often your brand appears across a defined set of category-relevant queries, not how often it appears in published coverage.
The competitive set can emerge from the model as well as from your methodology. Traditional SoV usually begins with a competitive set you define. AI answers can also surface competitors you did not include, revealing which brands the model associates with your category.
It operates at the query level. You might hold strong AI SoV on broad "best platform" queries and weak AI SoV on specific use-case queries. Topic-level AI SoV is usually more actionable than the aggregate.
Delve's AI Visibility tracks this AI-era share of voice, measuring how often your brand appears and is cited against competitors in AI answers across the major models. For the full framework covering AI visibility score, citation share, and narrative accuracy, see What Is PR Analytics?.
"What is the source coverage, and what is in the denominator?"
A share-of-voice number only means something once you know what the denominator includes. Understand the coverage universe and whether it matches where your relevant coverage actually appears.
"Can I define my own competitive set, or is it fixed?"
Your strategic competitive set may differ from the category definitions built into a platform. Being able to define which brands count as competitors is a prerequisite for numbers that mean anything.
"Can I segment by topic, channel, or time period?"
Overall SoV is rarely actionable. Topic-level and channel-level segmentation is where SoV turns from a reporting metric into a decision-making tool.
"How does it weight coverage, by outlet quality, audience, or raw counts?"
Depending on the reporting goal, weighting by outlet quality or audience can be more informative than raw counts. Know which method you are being quoted, and whether any reach figures are actual or estimated, before comparing across vendors or periods.
"Does it track AI answer-engine share of voice?"
For audiences researching through AI tools, traditional SoV is an increasingly incomplete picture. Ask for a specific answer on AI visibility tracking and what it includes.
Is share of voice the same in PR as in advertising?
No. In advertising, share of voice usually means your share of advertising activity or exposure in a category, often calculated from spend or impressions. In PR, it means your share of earned media coverage. Same logic, different activity.
Should I measure share of voice monthly or quarterly?
Both, for different purposes. Monthly tracking is useful for news cycles and campaign responsiveness. Quarterly views smooth out one-time spikes and are often better suited to strategic evaluation.
Can share of voice be gamed?
Yes. If press releases, syndications, or low-value pickups count toward qualifying coverage, high-volume publishing can inflate SoV without adding meaningful earned-media presence. Deduplication, consistent inclusion criteria, and outlet-quality filters make the metric harder to game.
What is the relationship between share of voice and market share?
Research linked to Binet and Field in advertising ties excess share of voice, SoV above market share, to brand growth. That should not transfer directly to earned media, where SoV indicates competitive visibility but does not predict market-share growth.
Share of voice is a useful starting point and a poor destination. It tells you how much of the conversation you occupy. It does not tell you whether that conversation is working for you, moving in the right direction, or reaching the audiences that matter.
The complete picture requires reading SoV alongside sentiment trajectory and narrative alignment, in What Is PR Analytics?, understanding the difference between SoV and narrative control, in Share of Voice vs. Share of Mentions, and extending the measurement into the AI answer engines where your audiences increasingly do their research.
