Share of Voice Calculator
Enter the counts you already have. Get share of voice for every brand in your set, excess share of voice against your market share, and a result card you can download. Nothing is fetched, you supply the numbers.
Result card
Your brand
Enter a value above zero for at least one brand.
All three figures were fitted on advertising spend share, not on any of the substitute proxies in this tool. Nielsen tested 123 brands across 30 categories in 2009 and offered 0.5 as a norm for fast moving consumer goods; the 0.6 and 0.7 come from the IPA Effectiveness Databank campaigns Binet and Field analysed for LinkedIn, published in 2020. Nielsen adds that variance across categories and brands was large, and a 2009 consumer goods figure is not a current cross-category constant. An indication, not a promise of growth.
What this does not tell you. Whether the metric above behaves like advertising spend, whether this category has a measurable total size, or when any effect would arrive. Voice moves in weeks and category share moves across renewal cycles, so read this against a market share figure from an earlier quarter.
Where this proxy breaks: three counters on one platform on one day gave the same vendor 1.0%, 4.7% and 5.3%, and the spelling of the query moved another vendor by more than half. Publish the query string, not just the vendor name.
Shares rounded to one decimal place by the largest remainder method, so the column totals exactly 100.0. Every number was supplied by the user, nothing was fetched.
Join the newsletter. Positioning, launch playbooks, persona updates - no spam, just value.
How to use the share of voice calculator
- Pick the voice proxy first. Advertising spend is the original definition and the only input the excess share of voice rule was fitted on. Everything else in the list is a substitute, so choosing here decides what your percentage is actually measuring.
- Fix the competitor set before you count anything. Write the list down and keep it for the year. Adding or dropping one vendor changes every share in the table, which means a share that moved because the set moved is not a result.
- Count every brand the same way, over the same window. One platform, one query string, one date range. Three counters run on one platform on one day returned 1.0 percent, 4.7 percent and 5.3 percent for the same vendor, so mixing methods silently invents a winner.
- Enter your brand, then up to five competitors. Commas, spaces, a leading currency symbol and the magnitude suffixes k, m and bn are all read, so $1.2m and 1,250,000 both land as the same figure when you paste straight from a spreadsheet or an ad panel. Anything the parser cannot read is rejected rather than turned into a smaller number: that row drops out of the denominator and a note under the card names the value it ignored. Rows with no value are left out too, not counted as zero.
- Add your market share only if you have a defensible one. That field is optional because in most private categories nobody publishes the denominator. Without it you still get every share of voice percentage, you just do not get ESOV.
- Read the indication block as a range, not a number. The growth constant is published at 0.5 (Nielsen, 2009), 0.7 and 0.6 (LinkedIn with Binet and Field, 2020) by two different research bases, so the card prints all three with the year and the scope attached to each. If the spread between them changes your decision, the decision was not supported by the rule.
- Download the PNG and keep the caveat attached. The card exports 1000 pixels wide at 2x device pixel ratio, so the file lands at 2000 pixels across. The proxy name, the failure mode of that proxy, the rounding method and the scope of the growth constant are all printed on the card, so the screenshot cannot be separated from the conditions it was true under.
Where the three ESOV growth constants come from
The card prints every published version of the rule instead of picking one, because the three figures come from two research bases, measured on different buyers, more than a decade apart. All three sources are linked here so you can read the wording for yourself before you quote the number in a deck.
The 0.5 is the oldest and the most repeated. Nielsen's August 2009 analysis Budgeting for the Upturn: Does Share of Voice Matter? found that "on average, a 10 point difference between SOV and SOM leads to 0.5% of extra market share growth" across 123 brands in 30 product categories, and recommended the 10 to 0.5 ratio as a norm for fast moving consumer goods brands. The same page records the limit on its own finding: variances across particular categories and brands were large, so a brand should measure its specific SOV to SOM relationship to get an accurate benchmark.
The 0.7 and the 0.6 are a single source published in January 2020, The B2B Marketer's Guide to the Share of Voice Rule on LinkedIn's marketing blog, drawn from the work Les Binet and Peter Field did with LinkedIn on the IPA Effectiveness Databank (the IPA's index of their key works). That page puts B2B at "an ESOV of 10% leads to market share growth of 0.7% per year" and B2C at 0.6, and works the arithmetic through a cloud services brand holding 10 percent market share while buying 20 percent of category advertising.
Two caveats travel with all three. Every one of them was fitted on share of advertising spend, which is the first option in the proxy list and none of the other six, so a mention share or an AI citation share run through the same multiplier is an extrapolation rather than a reading of the research. And the 0.5 is a 2009 consumer goods figure, well past the five-year age limit anything else on this site would have to clear. It stays on the card because it is the genuine provenance of the most widely repeated form of the rule, and because its date belongs next to the 2020 figures where a reader can weigh it.
What makes a good share of voice measurement
A good share of voice measurement is one somebody else can rebuild. That means the metric is named in the label rather than left as the bare phrase "share of voice", the competitor set and the query that produced the counts are published alongside the percentage, and the raw counts sit next to the share. A figure like 125 out of 2,350 survives an audit six weeks later in a way that 5.3 percent does not, because the second version hides which denominator produced it. Four widely cited pages publish a share of voice formula and they do not agree on what belongs in the bottom of that fraction, with one of them dividing by your own auction eligibility rather than by anything the category did, so a number is only comparable to a benchmark computed the same way.
The second half is timing. Share of voice is an input that moves in weeks, and market share is an outcome settled by contracts already signed, so comparing the two inside one quarter compares two different moments. Research by Professor John Dawes for the LinkedIn B2B Institute, summarised on the Ehrenberg-Bass Institute's own site, found that companies change providers of services such as banking, legal advice, software or telecoms roughly every five years, so "only 20% are in the market for those services in a given year and just 5% in a given quarter" (Ehrenberg-Bass Institute). Voice reaching the other 95 percent cannot convert into share this quarter, because those accounts have nothing to buy. The full argument, the primary-source trace of the ESOV constant, the six substitute proxies and a worked calculation on public data are all in the companion guide:
Read the full guide: Share of Voice vs Market Share: ESOV Without Ad Spend
Related templates and guides
Frequently asked questions
What is share of voice?
Share of voice is your brand's portion of the total attention in a category, expressed as a percentage. The original definition counts advertising spend: your category ad spend divided by total category ad spend. Syndicated ad-spend panels measure placements in TV, digital, social, audio, print, out of home and cinema, so a budget spent on conference sponsorship, documentation and developer relations does not enter the panel at all. Teams substitute another countable signal such as mentions, organic visibility, review velocity or AI citations, and the substitution changes what the number means.
How do you calculate share of voice?
Divide your brand's volume of one chosen metric by the category total for the same metric over the same window, then multiply by 100. This calculator does that for your brand and up to five competitors at once, so every share comes from the same denominator. Shares are rounded to one decimal place using the largest remainder method, which makes the column total exactly 100.0 rather than 99.9 or 100.1.
What is excess share of voice, and how does the calculator use it?
Excess share of voice, or ESOV, is share of voice minus share of market in percentage points. LinkedIn's 2020 guide to the rule works it as a cloud services example: a brand holding 10 percent market share while buying 20 percent of category advertising has a 10 percent ESOV, which that page pairs with 0.7 percent annual share growth. Enter your market share in the optional field and the calculator prints your ESOV signed, labels it as a growth-side or deficit position, and shows the indicated annual share change at all three published constants: 0.5 from Nielsen's 2009 fast-moving-consumer-goods analysis, and 0.7 for B2B and 0.6 for B2C from the 2020 LinkedIn write-up of the IPA Effectiveness Databank work with Binet and Field. All three source pages are linked in the section above these questions.
What data do I need before I start?
One number per brand, counted the same way for every brand, over the same window. That is it. You also need the competitor set fixed in advance, because adding or dropping one vendor changes every share in the table. Your own market share is optional and only feeds the ESOV block. Paste figures in whatever shape your spreadsheet or ad panel hands you: commas, spaces, a leading currency symbol and the magnitude suffixes k, m and bn are all read, so 1,250,000 and $1.2m are the same input. Anything else is rejected rather than salvaged, left out of the denominator, and named in a note under the card. The calculator fetches nothing and connects to no data source, so whatever you paste in is the whole input.
Which voice proxy should I pick?
Pick the one you can count consistently for every brand in the set, then name it next to the number forever. Advertising spend is the only proxy the ESOV constant was actually fitted on. Organic SERP share, review velocity, mentions, sponsorship slots and AI answer citations are all substitutes, and the calculator prints the specific failure mode of whichever one you choose on the result card so it travels with the screenshot.
Is this share of voice calculator free, and does it store my numbers?
It is free, with no account and no signup. Everything runs in your browser. Your inputs are saved to your own device with localStorage so they survive a refresh, and the Clear saved button deletes that copy. Nothing is sent to a server.
What does the output not tell me?
It does not tell you whether the metric you counted behaves like advertising spend, whether your category has a measurable total size, or when any effect would arrive. The ESOV growth figures are an indication drawn from two advertising-spend datasets, not a forecast for your brand, and the constant sits at 0.5 (Nielsen, 2009), 0.6 or 0.7 (LinkedIn with Binet and Field, 2020) depending on which corpus you quote. Nielsen's own page adds that variance across categories and brands in its study was large and that a brand should measure its own SOV to SOM relationship rather than lean on the norm. The 2009 date matters too: that figure is old, and it was fitted on fast-moving consumer goods, so it is a reference point rather than a current cross-category constant. Voice moves in weeks and category share moves across renewal cycles, so the two numbers are not from the same moment.
How often should I measure share of voice?
Quarterly is enough for most B2B categories, and the window matters more than the frequency. Because share of voice is a leading indicator, compare voice from an earlier quarter against market share for the quarter just closed rather than reading both from the same three months. Research by Professor John Dawes for the LinkedIn B2B Institute found that companies change providers of services such as banking, legal advice, software or telecoms roughly every five years, so only 20 percent are in the market for those services in a given year and just 5 percent in a given quarter. That figure was measured on those service categories rather than on every B2B purchase, but where your category behaves the same way, voice reaching the other 95 percent cannot show up in this quarter's share.
What format is the output, and what size is the file?
Two formats, both generated in your browser. Download PNG renders the result card to a PNG at 1000 CSS pixels wide and a device pixel ratio of 2, so the saved file is 2000 pixels across; the height varies with how many brands you entered and whether the ESOV block is showing. The file is named after your brand, for example northwind-labs-share-of-voice.png. Copy results puts a plain-text version on your clipboard: the ranked table, your share of voice, market share and ESOV, the indication at all three constants, the failure mode of the proxy you picked, and the rounding method. There is no PDF and no spreadsheet export.
What makes a great share of voice measurement?
One that somebody else can rebuild from what you published. Name the metric in the label instead of writing the bare phrase share of voice, publish the competitor set and the exact query or keyword list that produced the counts, and keep the raw counts next to the percentage, because 125 out of 2,350 survives an audit six weeks later in a way that 5.3 percent does not. Then hold all of it still: same platform, same query string, same date range, same set, quarter after quarter. A great measurement is boring and repeatable, and it is compared against a market share figure from an earlier quarter rather than the one it sits beside.