You get:
- “Our voice is friendly” (what does that mean?)
- no distinction between voice (consistent) and tone (varies by channel)
- no do’s and don’ts — so writers guess
- no channel-specific guidance (email vs. social vs. support)
- no “voice violation” warning — so off-brand writing slips through
But brand voice is not a vibe.
It is a set of repeatable choices.
- Voice pillars give writers guardrails
- Do’s and don’ts make the abstract concrete
- Tone adaptation recognizes that context matters
- Example sentences show, don’t tell
Without a voice matrix, your brand sounds different to every writer.
This framework forces AI to build voice guidelines that writers can actually use.
Assume the role of a brand voice architect who defines how brands speak across channels. Your task is to generate a brand voice matrix. Generate: 1. VOICE PILLARS (3 dimensions) - Each pillar: what the brand is (e.g., "Witty") - Each pillar: what the brand is not (e.g., "Not Sarcastic") 2. DO'S AND DON'TS FOR EACH PILLAR - 2-3 examples of language that fits - 2-3 examples of language that violates 3. TONE ADAPTATIONS BY CHANNEL - Social media: (e.g., "More playful, shorter sentences") - Email: (e.g., "Warm, conversational, longer") - Website: (e.g., "Confident, benefit-driven") - Customer support: (e.g., "Empathetic, solution-oriented") 4. EXAMPLE SENTENCES (one per channel) 5. VOICE VIOLATION WARNING - What would sound off-brand (specific examples) INPUTS: Brand Personality (3-5 adjectives): [E.G., "Confident, witty, helpful, direct"] Target Audience: [WHO ARE YOU TALKING TO?] Industry: [INSERT] Channel Mix (which channels you use): [SOCIAL MEDIA / EMAIL / WEBSITE / SUPPORT / OTHER] Competitor Voice (optional): [WHAT DO COMPETITORS SOUND LIKE?] RULES: - Each voice pillar must include a "not" (e.g., "Witty, not Sarcastic") - Tone adaptations must be different by channel (same voice, different tone) - Voice violation warning must include a specific example sentence that would be wrong - Example sentences must feel natural, not forced - Avoid "professional" as a pillar (it's table stakes)
- Share the voice matrix with every writer, marketer, and support agent.
- Keep a living document — update as you learn what works.
- Use the voice violation warning as a “test” for new copy.
- Channel tone can vary, but voice pillars should remain consistent.
- Audit your existing content against the matrix — you’ll find violations.
Brand Personality: Confident, witty, helpful, direct
Target Audience: Small business owners who are tired of marketing jargon
Industry: Marketing software
Channel Mix: Social media (Twitter/LinkedIn), email newsletter, website copy, customer support chat
Competitor Voice: Most competitors sound corporate, jargon-heavy, and cautious
This framework improves outcomes by forcing:
- voice pillars with “not” statements (clarity)
- do’s and don’ts (actionable examples)
- channel-specific tone (context matters)
- example sentences (show, don’t tell)
- voice violation warnings (what to avoid)
Great brand voice guidelines don’t describe the vibe — they show writers how to sound like you.
Build Better AI Systems
Subscribe for advanced prompt engineering, AI marketing tools, brand positioning frameworks, and practical strategies for advertisers and business owners.