Brand voice guide
Build a brand voice guide your team can actually use.
A brand voice guide is a practical rulebook for how your company sounds. The best version turns real content into clear traits, do-and-don't examples, channel rules, and AI prompts that keep every post recognizably yours.
What to include in a brand voice guide
A useful guide documents writing behavior. It tells a person or AI system how to sound in public, what to avoid, and how to decide when a draft is off-brand. It should be short enough to use while reviewing a post.
Audience
Who you speak to, what they already know, and what they need from you.
Voice traits
Three to five traits with concrete writing behavior.
Tone by channel
How the voice shifts across LinkedIn, Instagram, email, support, and product UI.
Vocabulary
Words to use, words to avoid, preferred spellings, acronyms, and formatting rules.
Before and after
Examples that show how generic copy becomes recognizably yours.
AI prompt rules
Instructions an AI tool should follow before drafting anything public.
Brand voice chart
A voice chart turns traits into rules. This is the section that makes a guide usable for writers, reviewers, and AI prompts.
| Trait | Means | Do | Avoid |
|---|---|---|---|
| Clear | The reader knows the point in the first sentence. | Say the concrete thing first. | Opening with abstract positioning or vague value props. |
| Useful | Every post gives the reader something to apply, decide, or understand. | Use examples, constraints, and next steps. | Stopping at advice that could fit any brand. |
| Human | The copy sounds like a person on the team, not a campaign committee. | Use natural phrasing and real customer context. | Over-polished AI copy with no point of view. |
| Consistent | The same brand is recognizable across social, email, product, and support. | Document repeatable tone, rhythm, vocabulary, and formatting rules. | Letting every channel invent its own voice. |
Brand voice template
Start with 10 to 20 real posts, emails, or founder notes. Mark the lines that sound right, remove the lines that could belong to any competitor, then convert the patterns into rules.
Copy this structure
- 1. Audience: who the brand is talking to and what they care about.
- 2. Voice traits: three to five traits with behavior and examples.
- 3. Channel rules: what changes on social, email, support, and product UI.
- 4. Words: phrases to use, phrases to avoid, and formatting preferences.
- 5. Review checklist: a short list reviewers can run before publishing.
Brand voice adjectives only work when they have boundaries
Searchers often look for a brand voice adjective list, but adjectives alone are too loose. Pair each one with what it is not and how it changes a draft.
Clear but not simplistic
Use plain language without removing useful detail.
Friendly but not casual everywhere
Sound approachable while keeping the promise precise.
Expert but not academic
Explain the mechanism, then make the next step obvious.
Confident but not hype-heavy
State what the product helps with and avoid impossible guarantees.
Practical but not dry
Give examples, constraints, and review rules the team can apply.
For worked examples, use the brand voice examples guide. For a reusable structure, use the brand voice template.
How to use a brand voice guide with AI
The mistake is asking AI to "write in our brand voice" without giving it a usable voice system. A good prompt includes the audience, the traits, the channel, the forbidden patterns, and examples of the target style.
- 1. Analyze real samples. Use your best existing posts to find repeated tone, sentence length, vocabulary, and calls to action.
- 2. Turn patterns into rules. Write instructions an assistant can follow, such as "use short active sentences" or "avoid hype words like revolutionary."
- 3. Add contrast. Include one good post and one generic rewrite so the model has boundaries.
- 4. Review before scheduling. AI can draft in the voice, but humans should approve anything public.
Fast path: paste a few posts into the free Brand Voice Analyzer. It extracts the first voice profile, which you can turn into prompt rules and review criteria.