Share of voice is the percentage of total conversation, coverage, or ad visibility in a category that belongs to your brand rather than a competitor. The formula is simple: divide your brand's mentions, impressions, or coverage by the combined total for the whole category, then multiply by 100. What makes the metric worth tracking closely isn't the formula. It's what the number predicts. Research from Les Binet, Peter Field, and the Ehrenberg-Bass Institute has repeatedly found that brands spending a higher share of voice than their current share of market tend to grow, while brands underspending their market share tend to shrink. This guide covers the formula in full, worked examples across search, PR, social, and paid channels, and the excess share of voice concept that turns a visibility number into a growth forecast.
- The formula never changes across channels: your metric divided by the category total, times 100. Only the metric being counted changes.
- A 10-point excess share of voice (share of voice minus share of market) correlates with roughly 0.7% annual market share growth in B2C and 0.6% in B2B, according to Binet, Field, and Ehrenberg-Bass Institute research.
- Five channels, five different metrics. Search uses ranking visibility, PPC uses impression share, PR uses weighted media mentions, social uses mention volume, and AI answer engines use citation rate.
- Share of voice is not the same as share of market. A brand can dominate the conversation and still trail on revenue, or the reverse, and the gap between the two is the actual signal worth watching.
- AI answer engine share of voice is the newest and least tracked channel. Most brands still measure search, social, and PR closely while having no idea how often ChatGPT or Perplexity names them versus a competitor.
What is share of voice?
Share of voice started life as a paid-media metric decades ago, measuring what percentage of total ad spend or airtime in a category belonged to one advertiser. A cereal brand buying 30% of the ad slots during a category's peak season had a 30% share of voice, regardless of how much cereal it actually sold. That distinction, visibility versus sales, is the whole reason the metric survived the shift to digital.
Today the term covers far more than ad spend. It applies to organic search rankings, social mention volume, press coverage, and now AI answer engine citations, anywhere a category gets discussed and a brand can own a bigger or smaller piece of that discussion. Brand monitoring is the practice that supplies the raw mention data share of voice is calculated from, and the two concepts get confused often enough that it's worth being precise: monitoring collects the mentions, share of voice is one specific metric calculated from what monitoring collects.
Share of voice always answers a comparative question, never an absolute one. Ten thousand mentions sounds like a lot until a competitor pulls twenty thousand in the same window. The raw number tells a team almost nothing on its own. The percentage against the category total turns mention volume into an actual competitive read.
The formula
The core formula holds across every channel:
A quick example makes it concrete. Say a brand pulls 500 social mentions in a month, and the full category, that brand plus its three closest competitors, pulls 10,000 mentions combined over the same window. 500 divided by 10,000 is 0.05. Multiply by 100 and the brand's social share of voice for that month is 5%.
The metric on either side of the division changes by channel, mentions, impressions, ranking visibility, ad spend, but the shape of the calculation never does. That consistency keeps share of voice comparable across a marketing stack that otherwise reports in five different units, impressions here, rankings there, mentions somewhere else.
Share of voice by channel
Five channels account for most of the share of voice tracking that actually happens, and each one swaps in a different metric for "mentions."
| Channel | What gets counted | Typical source |
|---|---|---|
| Search (SEO) | Ranking visibility across a keyword set, weighted by search volume and position | Ahrefs, Semrush |
| Paid search (PPC) | Impression share, the percentage of eligible auctions your ad actually appeared in | Google Ads |
| PR and media | Weighted press mentions, often adjusted for outlet reach and prominence | Media monitoring tools |
| Social | Raw mention volume across platforms, sometimes weighted by engagement | Social listening tools |
| AI answer engines | Citation rate across a fixed set of category queries in ChatGPT, Perplexity, and AI Overviews | AI visibility tools |
PR share of voice usually needs a weighting step the other channels can skip. A single paragraph in a niche trade blog and a full feature in a major outlet both count as "one mention" if a team isn't careful, and treating them equally flattens a real difference in reach. Press coverage tracking that accounts for publication reach gives a more honest PR share of voice number than a flat mention count.
Social share of voice runs into a different problem: platform mix. A brand that dominates Reddit conversation but barely registers on X has a real, uneven share of voice that a single blended number hides. Tracking Reddit specifically alongside the broader social number usually surfaces a gap most teams didn't know they had, since forum conversation skews toward exactly the kind of detailed comparison discussion that a blended mention count buries.
AI answer engine share of voice works differently from the other four. There's no fixed inventory of ad slots or search results pages, an AI model can name one brand, three brands, or none at all in a single answer, so the "total category mentions" denominator has to be built from a fixed query set run consistently over time. AI visibility tracking handles this by running the same category questions on a schedule and logging which brands get named each time, which is the closest equivalent to impression share this channel currently has. For more on how citation differs from ranking, see our GEO versus SEO breakdown.
Three worked examples
Formulas read differently once real numbers sit inside them. Here are three, each from a different channel.
Organic search. A site pulls 15,000 monthly organic visits across its target keyword set. The full competitive set, that site plus every competitor ranking for the same keywords, pulls 100,000 combined monthly visits. 15,000 divided by 100,000, times 100, gives a 15% search share of voice.
Paid search. A Google Ads account shows a 40% impression share on its core keyword set. That number is already a share of voice calculation; Google is running the division for you. It means the account's ads appeared in 40% of the auctions it was eligible to compete in, and the other 60% went to a mix of competitors and, sometimes, nobody at all if a competitor's budget ran out first.
PR and media. A brand's comms team lands 35 media placements in a quarter. The full category, again including every named competitor, generates 140 placements over the same three months. 35 divided by 140 is 0.25. That brand holds a 25% PR share of voice for the quarter, meaning one in four category stories mentioned it by name.
Notice that none of these numbers say anything about whether the coverage was positive. A brand pulling 25% PR share of voice during a product recall is not in a good position just because the formula produced a large number. Share of voice measures presence, not favorability, which is exactly why it needs to sit next to a sentiment number rather than replace one.
Excess share of voice and growth
Excess share of voice, ESOV for short, is share of voice minus share of market. A brand holding 20% share of voice against a 15% share of market has an ESOV of positive 5 points. A brand holding 15% share of voice against a 20% share of market has an ESOV of negative 5, meaning it is being outspent, relative to its size, by the rest of the category.
A 10-point excess share of voice correlates with roughly 0.7% annual market share growth in B2C categories and 0.6% in B2B, based on research from Les Binet, Peter Field, and the Ehrenberg-Bass Institute across decades of advertising effectiveness data.
The mechanism behind that correlation comes down to something the Ehrenberg-Bass Institute calls mental availability. Most category buyers aren't actively shopping most of the time. They're light or occasional purchasers who need to be reminded a brand exists at the exact moment they're finally ready to buy, and a brand outspending its market share simply reaches more of those buyers, more often, than a brand spending in line with or below its current size.
Small brands benefit from this dynamic more than large ones proportionally, which is the part most teams miss. A brand with 2% market share only needs to hold a modest share of voice above 2% to post a large ESOV in percentage-point terms, since the bar it's clearing is so low to start. A brand already sitting at 40% market share needs a genuinely large voice investment to move its ESOV by the same number of points. That asymmetry is why the rule shows up as a growth lever most often at the challenger-brand end of a category, not the incumbent end.
Improving your share of voice
The lever differs by channel, but the starting point is the same everywhere: know the current number before spending a dollar trying to move it.
- Search. Target keyword gaps where a competitor ranks and the brand doesn't, rather than trying to outrank an already-strong position. New visibility moves the percentage faster than incremental gains on keywords already won.
- Paid. Google Ads reports the specific reason share is being lost, budget or ad rank, directly in the impression share columns. A budget-capped account is a straightforward fix. A rank-capped account needs creative or landing page work first.
- PR. Original research and data-backed stories consistently outperform straight announcements for earning coverage, which is one reason data-driven pieces tend to punch above their weight on PR share of voice specifically.
- Social. Consistency beats intensity. Brands with the strongest social share of voice tend to show up steadily rather than in occasional bursts, since the metrics that actually matter reward sustained presence over campaign spikes.
- AI answer engines. Named entities, defined clearly, with attributed data points, get cited more often than pages optimized only for traditional ranking. This channel rewards content built to be extracted, not just found.
Tracking named competitors turns a share of voice program from a vanity dashboard into a working one. A number that only shows your own trend, with no competitor baseline next to it, cannot tell you whether a 10% jump is a real win or just the whole category growing at once. Agencies running this across several accounts at once standardize that comparison set per client rather than rebuilding the competitor list from scratch every time.
Common mistakes
- Reporting share of voice without a sentiment number next to it. A spike in mentions during a crisis reads as growth on a raw share of voice chart. It is the opposite.
- Comparing share of voice across inconsistent time windows. A monthly number and a quarterly number are not comparable, and blending them produces a trend line that means nothing.
- Leaving out a real competitor because tracking them is inconvenient. The category total has to include every brand actually competing for the same buyer, not just the ones with an easy API to pull from.
- Treating every channel's share of voice as interchangeable. A strong search share of voice and a weak social one are not offsetting facts. They are two separate problems that need two separate plans.
- Ignoring AI answer engines entirely. A brand can hold a commanding share of voice across search, PR, and social while barely registering when someone asks an AI assistant the same category question, and most reporting still has no line item for that gap.
Frequently asked questions
What's the difference between share of voice and share of market?
Share of voice measures visibility, mentions, coverage, or ad presence relative to the category total. Share of market measures actual sales or revenue relative to the category total. A brand can lead on one and trail on the other, and the gap between the two, excess share of voice, is often a better growth predictor than either number alone.
How often should share of voice be measured?
Monthly is the standard cadence for most channels, since it smooths out day-to-day noise while still catching a real shift within a quarter. PR and AI answer engine share of voice benefit from more frequent spot checks around known events, an earnings call, a product launch, a competitor's announcement, since those are exactly the moments the number tends to move fastest.
Can a small brand realistically compete on share of voice against a market leader?
Yes, and the excess share of voice math actually favors small brands here. A challenger brand needs a much smaller absolute spend increase to post a large ESOV than an incumbent does, since the incumbent's market share baseline is already high. This is one reason ESOV shows up most clearly as a growth lever at the smaller end of a category.
Does a higher share of voice always mean more sales?
Not directly, and not immediately. The research behind excess share of voice measures a correlation with market share growth over roughly a year, not an instant sales lift. A brand can also post a high share of voice from negative coverage, which drives visibility without driving purchase intent, so the number needs sentiment context before anyone treats it as a straightforward win.
What tools actually calculate share of voice automatically?
Ahrefs and Semrush handle search share of voice through their visibility and market share features, both built specifically around ranking data rather than raw mentions. Google Ads reports PPC share of voice natively as impression share, which is really just this same formula pre-calculated inside the platform's own reporting. PR, social, and AI answer engine share of voice are a different problem entirely, since none of those three have a single vendor sitting in the middle of every transaction the way Google does for paid search. For those channels, a dedicated brand monitoring platform is generally required, one that can pull mention data across sources and compare it against a named competitor set, rather than relying on any single vendor's internal auction or ranking data. Most teams end up running two or three tools side by side rather than finding one that covers everything.
Share of voice is a simple division that happens to predict something real. The formula takes five minutes to calculate once the mention data exists. Getting the category total right, including every real competitor, weighting PR mentions honestly, and now accounting for AI answer engines, is the part that actually takes work. Pull your own share of voice for one channel this week, compare it against your share of market, and see which direction the gap points before deciding where next quarter's budget goes.
Track share of voice across every channel that matters
Mentient calculates share of voice against named competitors across press, social, reviews, and AI answer engines, in one dashboard. Start free and see where you actually stand this week.
Start free trial
