B2B Share of Voice Counts Only When Buyers Know It's You

Each week, count searches for your name against the competitors you meet most, and your share of mentions where your buyers read. Add a weekly count of the idea you want to be known for. Once a month, read a sample of those mentions to see whether they say anything only you say.

Most guides stop at the count: your mentions divided by all the mentions in your market. The count is worth keeping. It moves week to week, and brand work rarely gets a figure leadership will accept on that clock.

The trouble is what the count leaves out. A brand that sounds exactly like its rivals can win share of voice while buyers still can't tell it apart.

Share of search is the weekly share of voice a B2B team can afford

The share of voice metric is your brand's share of the advertising and mentions in your market. The cheapest version you can run every week is share of search.

Les Binet, head of effectiveness at adam&eveDDB, set out the measure in a presentation published by the IPA. Share of search, he said, "equates to total searches for a specific brand, divided by the total searches for all brands in that category."

He also warned that share of search has limits and shouldn't be treated as the whole answer.

The searches are organic Google searches. The figures are free from Google Trends and can be pulled weekly.

To run it, list the competitors you meet most often when you're trying to win a deal. Pull the weekly search figures for each of their names and for yours. Divide your figure by the total for the whole list, your name included, and the result is your share of search for that week.

Share of voice earns a place in the report because it can move before market share

The reason to report share of voice is the comparison with share of market. The gap between the two is called extra share of voice, or ESOV. A brand whose share of voice is higher than its share of market has positive ESOV.

LinkedIn, which sells B2B advertising, makes the case for ESOV, drawing on IPA effectiveness data, in its guide to the share of voice rule.

The guide says that in B2B "an ESOV of 10% leads to market share growth of 0.7% per year," and that the rule "appears to have an even stronger effect in B2B than in B2C."

Binet's presentation shows how far share of search can run ahead of market share. He found share of search tracked market share in cars, energy and mobile handsets, and that "the lead time for Share of Search on Share of Market can be substantial, up to a year for cars."

A measure that can move ahead of sales gives the brand team something to show while sales catch up. Without one, brand building tends to lose budget to the weekly report.

Binet's share of search findings have further limits. Marketing Week reported a think tank study that found share of search represents 83% of a brand's market share, a ratio that "does vary by category and market." Its 30 case studies name no B2B categories, and its results are correlations, so they can't show that search drives market share.

A mention count can't tell whether buyers credit the mention to you

A mention count files every ad and article under the brand that paid for it or was named in it. Buyers do their own crediting. In one logo-blind test of cloud ads, they credited much of what they saw to Microsoft, the market's biggest name.

LinkedIn's B2B Institute tested whether people could tell lookalike B2B ads apart, in its sea of sameness research.

The institute showed people the colours and ads of five cloud brands with the logos hidden: IBM, Dell, Microsoft Azure, DigitalOcean and Qumulo. Only 5.75% matched the colours to the right brands. Dell's ads were correctly attributed by 2% of people, while 11% thought they were Microsoft's. DigitalOcean's fared no better, also at 2%.

The 2022 study, run with Distinctive BAT, found the cloud market's assets credited to Microsoft nearly twice as often as to any other brand. Haley Pierce, who wrote up the research, concluded that, without logos, it was "nearly impossible to distinguish one brand from another."

Like the share of voice guide, this research comes from LinkedIn. The write-up gives no sample size, and the 2022 study is proprietary. The test measured whether people could attribute colours and ads, a narrower thing than share of voice, so it shows the gap between a count and credit.

Run Dell's ads through a share of voice count and every one of them lands in Dell's column. More of the people who saw them put them in Microsoft's.

Pierce puts the cost plainly: "ads that look like your competitor are ads for your competitor." For a brand that looks like its market leader, a rising count can mean rising spend on the leader's behalf.

Our guide covers how buyers choose between vendors who look alike.

Count your name every week and read the mentions every month

Every week, record two counts for your name. The first is share of search against the competitors you meet most, built as above. The second is your share of mentions. Fix a list of the places your buyers read, such as the trade titles and newsletters in your market, and keep it the same each week.

Count the articles and posts there that name you, then the ones that name any competitor on your list. Divide yours by the total.

Then add a third count: mentions of the idea you want to be known for. Write that idea in a few plain words.

Each week, count how often the idea shows up in your market and how many of those mentions name you. A rising count for the idea with your name missing means the idea is spreading without you.

Once a month, pull a sample of the mentions and read them. Sort each one into two piles: mentions that say something only your company says, and mentions that say what any competitor could say. Both piles count toward share of voice. Track how big the first pile is from one month to the next.

A mention that names you but says what any rival could say works like the cloud test's lookalike ads. The count files it under you, and a buyer has nothing to remember it by.

The interchangeability test applies the same sort to your own copy.

For a share of voice example, set your weekly share of voice beside your market share to see what it is worth. LinkedIn's guide illustrates this with a cloud brand. Its figures are an example, and you can swap in your own.

First, take the brand's market share: 10%. Second, take its share of voice: 20%. Third, subtract market share from share of voice. That gap is the ESOV, here 10%.

Fourth, read off what that ESOV is worth. The guide's B2B rule says an ESOV of 10% leads to about 0.7% market share growth a year, so that's what the brand can expect on its current plan.

Now set a target. To grow 2% a year, the guide says the brand would need 40% share of voice. With market share still at 10%, that's an ESOV of 30%, which the guide puts at about 2.1% growth a year.

	Market share	Share of voice	ESOV	Expected growth a year
Current plan	10%	20%	10%	About 0.7%
2% growth target	10%	40%	30%	About 2.1%

The arithmetic tells you how much share of voice the brand would have to buy to reach the target. Where share of voice sits in the wider budget argument is covered in our guide to brand vs performance marketing.

The monthly read tells leadership what the weekly count is measuring

Won't a subjective monthly read give the CFO a reason to doubt the one number the brand team has? The worry is fair. A count is the brand figure leadership already accepts, and a judgement call attached to the count can look like a way to explain away a bad week.

So keep the count as the weekly number and put the monthly read beside it. The monthly read answers a separate question: whether the count is measuring your brand or your whole market.

If buyers credit your ads to the market leader, as they did in the cloud test, a rising count could mean you're paying to promote a rival. A finance lead would want to know that before approving next quarter's spend.

If you'd like to talk through what your share of voice is really counting, and how to tell whether the mentions are yours, get in touch with us.

If the first pile grows while the count holds steady, the same spend is earning more of the credit.