Brand Voice: How to Define, Document, and Enforce It

Brand voice is how your organisation sounds in writing. Not what you say — how you say it. Two companies can both write about content quality, but one sounds like a knowledgeable friend and the other sounds like a corporate brochure. The difference is brand voice.

Most teams skip brand voice or handle it poorly. They either don't define it ("we'll know it when we see it"), define it too vaguely ("be professional and friendly"), or define it perfectly but never enforce it (a gorgeous brand guide that nobody reads). The result is content that sounds like whoever happened to write it — not like the brand.

This guide covers all three phases: defining your voice, documenting it so writers can follow it, and enforcing it so consistency isn't optional.

Quick answer: Define brand voice with 3-5 personality traits on a spectrum (what you are vs. what you aren't). Document it with specific rules, vocabulary lists, and before/after examples in a 5-page writer's guide. Enforce it with AI-powered brand voice review — a reviewer with your guidelines as a knowledge base, scoring criteria for tone/terminology/consistency, and a quality gate that content must pass before publishing.

Phase 1: Define Your Brand Voice

Start With Personality Traits

Your brand voice is a personality expressed in writing. Define 3-5 personality traits that describe how your brand communicates.

The key: define each trait on a spectrum — what you are on one end and what you're not on the other. This eliminates ambiguity.

Example:

Trait	We Are...	We Aren't...
Confident	We state things directly. We use data and evidence. We take positions.	We don't hedge, qualify everything, or hide behind "might" and "perhaps." We also don't overstate — we prove, not claim.
Helpful	We explain how and why. We give actionable advice. We anticipate follow-up questions.	We don't just tell people what to do without explaining the reasoning. We also don't over-explain simple concepts.
Human	We write like a smart colleague. We use natural language, contractions, and the occasional question.	We don't use corporate jargon, buzzwords, or unnecessarily complex vocabulary. We also don't try to be funny if it doesn't serve the content.
Direct	We get to the point. First sentence of every section delivers value. Short paragraphs. Clear structure.	We don't ramble, over-qualify, or bury the point under three paragraphs of context.

Identify Your Anti-Voice

Sometimes it's easier to define what you're NOT:

We don't sound like a press release
We don't sound like a textbook
We don't sound like a salesperson at a trade show
We don't sound like ChatGPT's default output

Find 3-5 pieces of content (yours or competitors') that represent what you DON'T want to sound like. Analyse why — the specific phrases, constructions, and patterns that make them wrong. These become your "avoid" rules.

Test With Real Content

Take your personality traits and apply them to a recent piece of content. Does the piece match? Where does it deviate? The gaps between your defined voice and your actual content reveal where your guidelines need to be most specific.

Phase 2: Document Your Brand Voice

A defined voice that isn't documented is just an idea in someone's head. Documentation makes it transferable — new writers, freelancers, agencies, and AI tools can all reference the same standard.

The Writer's Quick Guide (5 Pages Max)

Create a concise reference document that writers use while writing. This is NOT the comprehensive brand bible. It's the practical extract that covers 90% of daily writing decisions.

Page 1: Personality Traits The spectrum table from Phase 1. Three examples per trait showing right vs. wrong.

Page 2: Vocabulary Rules Three lists: preferred terms (always use), banned terms (never use), and product terminology (exact names and capitalisation).

Page 3: Writing Rules Concrete rules for sentence length, paragraph length, active voice, perspective, contractions, numbers, headings, lists, and opening sentences.

Page 4: Channel Tone Matrix How each personality trait expresses differently by channel (blog, email, social, product, support).

Page 5: Before/After Examples 6-10 pairs showing common mistakes and their corrections. Cover: generic AI openings, hedging, jargon, passive voice, feature-focused vs. benefit-focused, and tone inconsistency.

→ Download: Brand Voice Guidelines Template

What Makes Good Documentation

Specific over aspirational. "Never start a blog post with 'In today's...'" is better than "write engaging openings."

Rules over adjectives. "Use active voice for all value propositions" is better than "sound confident."

Examples over descriptions. A before/after pair communicates more than a paragraph of explanation.

Short over comprehensive. A 5-page guide that gets read beats a 40-page guide that gets skimmed once.

Distribute and Socialise

The guide should be:

Accessible in the team's daily workspace (not buried in a drive)
Referenced during onboarding (every new writer reads it first)
Cited in feedback (editors say "see page 3 of the voice guide" not just "fix the tone")
Reviewed and updated quarterly

Phase 3: Enforce Your Brand Voice

Definition + documentation isn't enough. Without enforcement, guidelines are suggestions. Enforcement means every piece of content is checked against the documented voice before publishing.

Set Up AI-Powered Brand Voice Review

Upload your writer's quick guide as a knowledge base — the AI reviewer references your specific guidelines when scoring and giving feedback
Create a brand voice reviewer with five weighted criteria:
- Tone Alignment (30%) — does the content match your personality traits?
- Terminology Compliance (25%) — preferred terms used, banned terms absent?
- Consistency (20%) — same voice throughout, no shifts?
- Channel Appropriateness (15%) — tone matches the target channel?
- Personality Expression (10%) — brand personality is evident?
Set a quality gate — minimum score for brand voice approval (start at 70, raise to 78+ over time)
Make it mandatory — content must pass the brand voice gate before it advances to human review

The Enforcement Workflow

Writer creates content
        ↓
Submit to brand voice reviewer
        ↓
Score + specific feedback referencing brand guidelines
        ↓
Below gate? → Writer fixes cited issues → re-submit
Above gate? → Advances to human review
        ↓
Human reviewer checks strategic fit and creative quality
        ↓
Publish

What Changes When You Enforce

Week 1-2: Scores are lower than expected. Writers are surprised by how many brand voice issues the reviewer catches. This is normal — it's catching the drift that was invisible before.

Week 3-4: First-draft scores improve noticeably. Writers are adapting to the criteria. Common issues (banned terms, passive voice, tone shifts) decrease as writers self-correct during drafting.

Month 2-3: Brand voice scores stabilise at 75-85. First-draft pass rate exceeds 50%. Writers have internalised the core rules. Feedback shifts from basic issues (terminology, structure) to nuanced issues (personality expression, channel appropriateness).

Month 3+: Brand voice is no longer a problem. It's a system. New writers ramp up in weeks instead of months because they get immediate, specific feedback from day one.

Maintaining Your Brand Voice Over Time

Quarterly Review

Every quarter, review:

Are the personality traits still accurate? (Brands evolve.)
Are there new banned terms to add? (New trends bring new clichés.)
Are there new channels that need tone guidelines?
Are the before/after examples still relevant?
What feedback do editors give that isn't captured in the guidelines?

Score Trend Monitoring

Track brand voice scores weekly (initially) then monthly:

Team average — should trend upward for the first 3 months, then stabilise at 80+
Per-writer trends — identify who needs coaching
Per-channel trends — identify channels where voice is drifting
New writer ramp-up — how quickly do new writers reach target scores?

Guideline Evolution

Brand voice isn't static. As your company grows, enters new markets, or shifts positioning, the voice should evolve intentionally. When it does:

Update the writer's quick guide
Update the knowledge base
Announce the change to the team with specific examples
Monitor score trends to ensure the new voice is adopted

→ Get started: Create a brand voice reviewer

→ Related: