AI Visibility

Share of Voice: What It Is, How to Calculate It, and Why AI Answers Now Count

Nipuna Jayasekara 11 min read

Share of voice is the percentage of total market conversation your brand owns against competitors in a given channel, like paid search, social, or PR. To calculate share of voice, divide your brand’s metric by the combined total for every brand in that space, then multiply by 100. Simple math. The hard part is deciding what to count, and where.

Most guides stop at the formula. This one goes further, because the biggest channel for brand conversation in 2026 isn’t one the classic tools even measure. More on that below. First, the basics.

What is share of voice?

Share of voice (SOV) is the slice of total market activity your brand captures in a channel, measured against every competitor in the same space. It started as an advertising metric, your ad spend versus total category spend. Today it stretches across organic search, social media, and PR mentions. A higher share of voice means more of the conversation is about you.

Think of it as a scoreboard for attention. If a hundred people are talking about running shoes and eighteen mention your brand, your share of voice is 18%. It doesn’t tell you whether they liked you, only that you showed up. That’s why smart teams pair it with sentiment.

Share of voice is the measurement layer that sits on top of both search and answer engines. If you’ve read our take on GEO vs SEO, SOV is how you keep score across all of it.

The share of voice formula (with a worked example)

The share of voice formula is: (your brand’s metric ÷ total market metric) × 100. The metric changes by channel, ad impressions, search clicks, social mentions, but the shape never does. Your slice, divided by the whole pie, times a hundred.

Here’s a worked example using brand mentions. Say five brands compete in your category, and over one month the tracked mentions break down like this:

BrandMentionsShare of voice
Your brand48024%
Competitor A62031%
Competitor B40020%
Competitor C30015%
Competitor D20010%
Total2,000100%

Your 480 mentions divided by the 2,000 total, times 100, lands at 24%. Competitor A leads at 31%. That gap is your target for next quarter. Notice the shares add up to 100, which is the quick sanity check that you counted every competitor and didn’t double-count anyone.

How to calculate share of voice, step by step

To calculate share of voice, pick one channel, count your brand’s activity, count the total across all competitors in that channel, then divide and multiply by 100. The method is identical everywhere. What changes is the input, and getting the input right is where most people slip.

Work through it in five steps:

  1. Define the market. List every competitor you want to measure against. Leave one out and your denominator is wrong.
  2. Pick one channel and one metric. Don’t mix paid impressions with social mentions in the same number.
  3. Set a fixed time window. A week, a month, a quarter, just keep it consistent so you can compare later.
  4. Pull the numbers. Your metric, then the same metric for every competitor on the list.
  5. Divide your figure by the combined total and multiply by 100.

The inputs shift by channel. Here’s how.

Paid search share of voice

Use impression share. Google Ads reports it directly as the percentage of impressions your ads got out of the total they were eligible for. No manual counting needed. If your impression share is 45%, that’s your paid share of voice for those keywords, and Google even tells you how much you lost to budget versus ad rank.

Organic / SEO share of voice

Weight your keyword rankings by search volume and click-through rate, then compare against competitors ranking for the same terms. A number-one spot on a high-volume keyword is worth far more than a number-eight spot on a rare one. Tools like Ahrefs and Semrush estimate this so you don’t have to build the spreadsheet yourself.

Social media share of voice

Count brand mentions across the platforms that matter to you, yours and every competitor’s, over the same window. This is the messiest channel, because mentions include misspellings, hashtags, and handles. A social listening tool does the collecting. Your job is defining what counts as a mention before you start.

PR and media share of voice

Tally press mentions, often weighted by outlet reach or authority. A feature in a national title outweighs a link in a small trade blog. PR teams usually pull this from a media monitoring database rather than counting by hand, and the honest ones weight for prominence, not just volume.

How to measure share of voice across channels

To measure share of voice, you need a data source for each channel and a consistent counting rule across all of them. No single dashboard captures everything, so most teams stitch together two or three tools and reconcile the numbers monthly. The reconciliation is the annoying part.

The first time I tried to line up a social listening tool’s SoV figure with a client’s PR agency’s number, they were off by roughly 30 points. Same brand, same month. They’d counted different mentions. That’s the trap with cross-channel SOV, the tools aren’t wrong, they’re answering different questions.

Here’s where the numbers usually come from:

  • Paid: Google Ads and Microsoft Advertising, straight from the impression share report.
  • SEO: Ahrefs or Semrush, using rank tracking weighted by volume.
  • Social: Brandwatch, Sprout Social, Meltwater, or Talkwalker for mention counts.
  • PR: Cision or Muck Rack for media coverage.

Pick your tools to measure share of voice based on the channels you actually compete in. A local service business barely needs PR monitoring. A challenger SaaS brand lives and dies by search. Match the stack to the fight.

What’s a good share of voice percentage?

There’s no universal “good” share of voice, because it depends on how many brands split the market. As a rough benchmark, if five brands compete evenly, 20% is par, so anything above that means you’re beating the field. What matters more than the absolute number is the direction, a rising share of voice tends to predict rising market share.

Share of voice vs share of market

Share of voice is how much of the conversation you own. Share of market is how much of the actual sales you own. The gap between them predicts growth. When your share of voice runs ahead of your share of market, something marketers call excess share of voice, you tend to gain ground over time.

This isn’t a hunch. Les Binet, Head of Effectiveness at adam&eveDDB, and marketing consultant Peter Field analyzed close to 1,000 campaigns in the IPA Databank and found that for every 10 percentage points of excess share of voice a brand holds, it can expect roughly 0.5% of market share growth per year. Nielsen’s FMCG work adds a wrinkle, brands already above 10% market share get about 2.5 times the return on that excess. Big brands compound faster. The takeaway is plain: talk louder than your sales suggest you should, and sales tend to follow.

The channel most brands miss: share of voice in AI answers

Every framework above assumes the conversation happens somewhere you can see it. Ads, search results, social feeds, press. But a growing share of your customers now ask ChatGPT, Gemini, Claude, Perplexity, or Google’s AI Overviews instead, and get an answer that names two or three brands and skips the rest. Your rank tracker can’t see this. Neither can your social listening tool.

The scale is already large. Around 68% of U.S. Google searches ended without a single click in early 2026, according to SparkToro and Datos data reported by Search Engine Land. AI Overviews now show on more than 20% of searches, and when they appear, click-through to the pages below drops by nearly 60% (Semrush). The answer is the destination now. If an AI names your competitor and not you, you lost that customer before they ever reached a website.

And here’s the part the old tools miss entirely: AI answers aren’t fixed. I’ve watched brands celebrate a single ChatGPT screenshot where they got named, then panic when the same question named a rival an hour later. One check proves nothing. Share of voice in AI answers is a frequency game, measured across many runs, or it’s noise.

How to measure your AI share of voice

To measure AI share of voice, run the same set of buyer questions across every major AI engine, many times, and count how often each brand gets named against the total. It’s the classic SOV formula, your mentions over total mentions, applied to answers instead of ads. The only way to do it reliably is to automate the runs, because doing it by hand doesn’t scale past a prompt or two.

That’s the job MentionsFlow was built for. It runs your chosen prompts across ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews on a schedule, then reports your share of voice against named competitors on the identical prompts, a Visibility Score from 0 to 100, and a citation leaderboard showing which sources the models pull from. When we run a client’s core prompt 25 times across ChatGPT and Perplexity, we routinely see three or four different “best tool” answers. That spread is exactly why a single check tells you nothing, and why frequency is the whole measurement.

If you’re comparing options, our rundown of the 10 best AI brand visibility tools lays out the field honestly, and there’s a deeper piece on brand tracking software if you want the broader category. Or skip the reading and see it on your own brand. MentionsFlow is an AI visibility tool with plans from $49/mo, and you can start a 14-day free trial with no card.

How to increase your share of voice

To increase share of voice, you either produce more of the content and coverage your competitors are winning, or invest to outspend them where it counts. There’s no shortcut around showing up more often than the other brands. But the smart version is targeted, not just louder.

Four moves that actually shift the number:

  • Publish for the questions buyers ask, not the keywords you wish they searched. Answer-first content earns citations in both search and AI.
  • Spend where you’re under-indexed. Excess share of voice only works if you aim it at gaps, not at channels you already dominate.
  • Earn citations from sources the models trust. In AI answers, being quoted by a high-authority page compounds, cited today means likelier cited tomorrow.
  • Stay consistent. Share of voice rewards presence over time, not one big campaign that fades.

Measure first, then move. You can’t grow a number you aren’t tracking, and in 2026 that increasingly means tracking the answer box, not just the search results.

Frequently asked questions

What is share of voice in simple terms?

Share of voice is the share of a market’s total conversation your brand owns compared to competitors. If a hundred people discuss your category and twenty mention you, your share of voice is 20%. It measures how visible you are, across ads, search, social, or AI answers, against everyone you compete with.

What is the formula for share of voice?

The share of voice formula is (your brand’s metric ÷ total market metric) × 100. The metric depends on the channel: ad impressions for paid search, mention counts for social, press hits for PR. Divide your figure by the combined total for all brands, then multiply by 100 to get a percentage.

What is a good share of voice percentage?

There’s no fixed target, because it depends on how many brands split the market. If five compete evenly, 20% is par, so higher means you’re ahead. The trend matters more than the number, a share of voice that climbs quarter over quarter usually signals rising market share.

What’s the difference between share of voice and share of market?

Share of voice is how much of the conversation you own. Share of market is how much of the actual sales you own. When your share of voice sits above your share of market, called excess share of voice, research links it to future market share growth, roughly 0.5% a year for every 10 points of excess.

How do you measure share of voice in ChatGPT and other AI answers?

Run the same buyer questions across ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews many times, then count how often each brand is named against the total. Because AI answers change between runs, one check isn’t enough. Tools like MentionsFlow automate the runs and report your share against named competitors.

What tools measure share of voice?

For paid, Google Ads reports impression share directly. For SEO, Ahrefs and Semrush estimate ranking share. For social, Brandwatch, Sprout Social, and Meltwater count mentions. For PR, Cision and Muck Rack track coverage. For share of voice inside AI answers, MentionsFlow measures it across five engines.


See your AI share of voice on your own brand. Start a 14-day free trial of MentionsFlow, no card required, and track how often ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews name you versus your competitors. Prefer to look before you sign up? Browse a sample report first.

← Back to blog

Keep reading