Reading a Share-of-Voice Report Before You Trust It
Objective: Given a weekly Share of Voice export across a 5-brand competitive set broken out by audience segment, find where the brand is actually strong even though the blended overall number looks flat.
You're the marketing analyst at TAC Security, benchmarking Share of Voice against four competitors after a product launch, and the headline number hasn't moved.
Don't stop at the blended SOV number. Segment it by audience to find where the launch actually changed the conversation.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free and requires no setup beyond a query per brand
Free, and a simple pivot is enough to split a 5-brand mention set by segment
Paid upgrades (optional, faster/deeper)
Enterprise-grade query syntax and larger source coverage once mention volume outgrows manual tagging
No access? Google Alerts plus manual tagging in Google Sheets covers the same workflow at lower mention volume
The process
1 step
Step 01 of 01
The lesson says to benchmark SOV by audience segment, because a brand can be strong in one segment's conversation and invisible in another while the blended total looks unchanged.
Overall SOV moved from 11% to 12% this month, barely worth reporting. Segmented by audience, SOV in enterprise-security discussions moved from 9% to 21%, while SOV in SMB-security discussions dropped from 13% to 8%. What does the blended number hide?
Procedure
- Pull the week's mentions for your brand and the 4-competitor set
- Tag each mention's audience segment (enterprise vs. SMB) based on the source and discussion context
- Calculate SOV separately within each segment, not just overall
- Compare segment-level trends to the blended trend to see what's actually driving or masking the overall number
SOV by segment, this month vs. last Overall: 11% -> 12% (looks flat) Enterprise-security: 9% -> 21% (real launch win) SMB-security: 13% -> 8% (quietly losing ground)
Healthy
Reports both segment trends, flags the SMB decline as something to act on even though overall SOV rose.
Unhealthy
Reports only the blended 11%->12% figure and calls the launch a modest success.
What this means
A blended SOV number can hide a real win and a real loss happening at the same time, canceling each other out in the topline metric.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| SOV trend looks flat month over month despite a major launch | Re-run SOV segmented by audience before concluding the launch had no effect | 30 min |
Final deliverable
A segmented SOV table showing where the brand actually gained or lost ground, alongside the blended figure.
See a reference example
Freshworks, SOV by Segment (excerpt) Overall SOV: 14% -> 15% Developer-tools segment: 6% -> 17% IT-ops segment: 19% -> 14% Takeaway: the launch resonated with developers but lost ground in IT-ops; the blended number hid both.
Success criteria
You're done when you can:
- Calculates SOV separately for at least two audience segments, not just the blended total
- Identifies a segment-level trend the blended number obscures