Measuring Share of Voice and Sentiment Lift From Influencer Campaigns

What share of voice actually measures

Share of voice (SOV) is the share of the total conversation in a category that your brand owns, compared to competitors. Per Brandwatch's own definition, the formula is: SOV = (your brand's mentions ÷ total market mentions) × 100. If your brand gets 2,500 social mentions and competitors get 10,000 combined, your SOV is 2,500 ÷ 12,500 × 100 ≈ 20% of the total 12,500 mentions. SOV can be tracked across several channels — social, organic search, paid advertising and PR/media — but Brandwatch itself warns against averaging SOV across channels: 40% social SOV and 5% search SOV tell two different stories, not one blended 22.5% number.

Sentiment: the other half of the picture

SOV only tells you how much people are talking about your brand — not whether it's a good thing. That's where sentiment analysis comes in. Per Sprout Social's own description, sentiment analysis is an AI technique that classifies text as positive, negative or neutral; Sprout's own tool uses a deep neural network that calculates a probability for each of the three labels and assigns whichever scores highest. The method has known limits Sprout itself flags: sarcasm, slang and backhanded phrasing ("not so bad") can mislead even good models, and idioms and regional tone rarely translate cleanly across markets — a reason to analyze Danish-language content natively rather than through machine translation. Sprout also states plainly that AI sentiment doesn't replace human judgment for sensitive cases — it provides consistency at scale, not a final verdict.

Why a raw volume number can mislead you

The central problem: a rise in SOV can just as easily come from a PR crisis as from a successful campaign. Brandwatch itself puts it this way — combining SOV with sentiment analysis reveals "not just how much people talk about your brand, but whether that conversation is positive or negative." A coordinated influencer campaign can genuinely create a temporary wave of mentions and a real SOV spike — but if sentiment turns negative at the same time (because the content reads as pushy advertising, for example), that spike isn't a win.

SOV and sentiment are also an observational method — they count what's already being said, with no control group. That's a different approach from a brand lift study, which actively surveys an exposed group against a control group and can therefore support a causal claim. SOV/sentiment is cheaper and runs continuously, but on its own it doesn't prove your campaign moved the number — only that the conversation changed during that period.

Not the same as a social listening tool inside a platform

As what is an influencer marketing platform already notes, a social listening tool is its own category, separate from an influencer marketing platform's four core functions (discovery, campaign management, payment, tracking). SOV and sentiment are typically measured through a dedicated social listening tool (Brandwatch, Sprout Social and similar) — not through the platform you use to run the influencer collaboration itself. The two tool types complement each other but solve different jobs.

Three measurement methods side by side

Decision framework: when each measurement earns its keep

IF you want to track your position in a competitive category over time → SOV, measured continuously, is the right tool — treat it as a thermometer, not a campaign report.

IF you need to prove to leadership that a specific campaign changed people's perception → a brand lift study is the method that can actually support a causal claim.

IF you just need to know how many people saw and engaged with a specific influencer campaign → standard KPIs as covered in influencer campaign KPIs by objective are faster and cheaper.

IF tracking actual sales is the goal, not mentions or awareness → see influencer marketing attribution explained or the proxy signals covered in how to measure influencer marketing when you can't track everything instead.

Worked example: reading a SOV/sentiment shift

The figures below are hypothetical and illustrate only the calculation principle — this is not a real Make Influence customer case.

A skincare brand runs a two-week influencer campaign. The week before the campaign, the brand's social listening tool logs 420 mentions of the brand against 2,800 combined mentions of its three biggest competitors — a SOV of 420 ÷ (420+2,800) × 100 ≈ 13.0%. During the campaign week, the brand's mentions rise to 960, while competitors' combined mentions hold steady at 2,750 — SOV rises to 960 ÷ (960+2,750) × 100 ≈ 25.9%, nearly doubling.

But the sentiment split shifts too: before the campaign, 68% of the brand's mentions were positive, 24% neutral and 8% negative. During the campaign week, the positive share drops to 54%, while negative mentions rise to 19% — several users comment that the campaign "feels like too much advertising at once." SOV more than doubled, but the positive share fell by 14 percentage points. The conclusion isn't unambiguously positive: the campaign generated attention, but a real share of it was critical — a signal to adjust tone going forward, not just push more volume.

Common mistakes

• Celebrating a SOV rise without checking sentiment. A spike in mentions can easily be negative.
• Averaging SOV across channels. Social SOV and search SOV measure different things and should be reported separately, not blended into one number.
• Treating SOV/sentiment as proof of causation. The method shows a coincidence in timing, not proof the campaign caused the change — use a brand lift study if that's what needs documenting.
• Trusting AI sentiment blindly on sensitive cases. Sarcasm and ambiguous comments can be misclassified — check the biggest swings manually.

Make Influence's operational perspective

In Make Influence's experience, SOV and sentiment are often either ignored entirely or misread as a direct campaign KPI, which they aren't. We recommend using them as an ongoing background measure of the brand's position in the category — something checked quarterly, not per campaign — and reserving an actual brand lift study or standard performance KPIs to document a single campaign's result. A rising SOV number is interesting, but it's the sentiment line that tells you whether it's worth celebrating.

FAQ

Is share of voice the same as reach?

No. Reach is how many unique people saw your own content. SOV is your share of the total conversation about the category — including mentions you didn't create or pay for yourself.

Can SOV be measured without a paid social listening tool?

Partially. Free searches on specific hashtags or brand names can give a rough sense, but a dedicated tool such as Brandwatch or Sprout Social is needed for a reliable, bounded market number and automatic sentiment classification.

How often should SOV be measured?

Continuously — SOV is a temperature check on a category over time, not a one-off number. Most brands that use it seriously check it monthly or quarterly and look for trends, not single days.

Can a negative sentiment spike still be a good thing?

Rarely on its own, but it can point to a real product or communication issue worth knowing about early — treat it as a warning signal, not only a success metric.

Do SOV and sentiment replace a brand lift study?

No. They measure something fundamentally different: SOV/sentiment observes the public conversation with no control group, while a brand lift study actively tests an exposed group against a control group and can therefore support a causal claim that SOV/sentiment cannot. See brand lift studies explained.