Most brand voice guidelines are written by brand teams and applied to marketing copy by marketing teams. The disconnect produces emails that read 'on brand' and don't convert.
The fix isn't to abandon brand voice. It's to make brand voice a constraint, not the goal.
Three failure modes we see often:
The structural issue: brand voice is treated as the goal of the email rather than as a constraint within which the email achieves its actual goal (conversion).
A useful reframe: every email has one goal. Brand voice is one of many constraints on how the email pursues that goal. The list of constraints typically includes:
Brand voice is one constraint among five. Treating it as the dominant constraint over conversion is the mistake.
Move 1: Write to the goal first; voice-edit second.
Draft the email with one job in mind. Get the structure right. Then go back through and tune the voice without changing the structure. If a brand-voice edit kills the structural move, revert the edit. The hierarchy: structure beats voice.
Move 2: Build the brand voice descriptor from emails that actually converted.
Most brand voice guides are written from aspirations ('we want to sound bold and authentic'). Better: extract voice from your highest-converting past emails. What patterns appear in the ones that worked? Those patterns are your brand voice. Aspirations that don't appear in converting work shouldn't be in the guide.
Our brand voice extraction tool does this automatically — point it at your 5-8 best emails, it pulls the descriptor.
Move 3: Treat brand voice as a lint, not a rewrite.
Lint the email against the brand voice descriptor after the structure is right. Flag words that conflict with the descriptor; suggest alternatives; don't force a wholesale rewrite. The dominant emotion you want during voice-editing: 'small tweaks, not major changes.'
A worked example. Brand voice says: 'conversational, opinionated, sentence-rhythm varied.'
First draft (structure-first):
Hi Sara,
>
I'm not pitching anything. I just want to know — if you've stopped using Northstar, what's the reason?
>
One sentence is enough. "Too complicated." "Don't have time." "Doesn't fit my workflow." All useful.
>
If you're still using it — also useful to know what you're using it for.
This draft is structurally sound (one job: get a reply with churn signal). It also happens to be on-brand (conversational, opinionated, varied rhythm). No further voice editing needed.
If the brand voice were 'formal, structured, authoritative,' a different draft would be needed:
Sara,
>
A note: we'd value your perspective on your recent experience with Northstar.
>
If you've stopped using the platform, three categories of reason cover the typical answers we hear: complexity, time, or workflow fit. A single-sentence reply indicating which (or another category) would be useful.
>
Equally — if you're still using the platform, hearing about your specific use case would help us serve you better.
Same structure, different voice. Both convert. The mistake is when the voice editing damages the structure to fit the descriptor.
The pattern we see most often when brand voice damages conversion:
The brand team got what they wanted (softer voice); the company lost conversion. This kind of voice editing should be reversed.
A healthier collaboration: brand team defines the voice descriptor; marketing keeps editorial control over individual sends; weekly review of high-converting and low-converting emails together; voice descriptor updated based on what actually works.
For any company with a brand voice guide:
For more on brand voice extraction and the rubric we use, our brand voice guide has the full framework. And if you want to extract a voice descriptor from your own past emails, try our brand voice tool.
Free anonymous run, no signup.
Try the editor