Marketing AI Skill

Brand Voice Analyzer

Maintain consistent brand voice across all marketing content by analyzing tone, vocabulary, and style patterns. Use when enforcing brand voice consistency, analyzing content tone, creating brand voice guidelines, checking messaging alignment, or training AI...

Brand Voice Analyzer & Enforcer

Ensure every piece of content sounds unmistakably like your brand — across all channels, teams, and formats.

Workflow

  1. Define brand voice framework: personality traits, tone spectrum, vocabulary rules, messaging pillars.
  2. Collect brand voice examples: best-performing content, approved messaging, competitor differentiation.
  3. Build brand voice profile: AI-analyzed patterns in tone, vocabulary, sentence structure, and style.
  4. Create voice guidelines document: rules, examples, forbidden patterns, channel-specific adaptations.
  5. Deploy voice checking: scan new content against brand voice profile for consistency scoring.
  6. Generate voice-aligned alternatives: suggest rewrites that match brand voice when deviations detected.
  7. Train content team: workshops, reference guides, automated feedback on draft content.
  8. Measure voice consistency: track scoring over time across channels and content creators.
  9. Evolve voice guidelines: update based on brand evolution, audience feedback, and performance data.

Brand Voice Framework

BRAND VOICE PERSONALITY MATRIX
================================

THE 6 BRAND PERSONALITY DIMENSIONS (Keller & Kaynard Framework):

  1. SINCERITY:
     → Traits: Honest, wholesome, practical, hardworking, warm
     → Examples: John Lewis, Disney, Dove, State Farm
     → Tone: Friendly, trustworthy, genuine
     → Vocabulary: Simple, relatable, story-driven
     → Avoid: Arrogance, hype, pretension

  2. EXCITEMENT:
     → Traits: Energetic, bold, spirited, imaginative, trendy
     → Examples: Red Bull, Nike, Apple, Spotify
     → Tone: Dynamic, enthusiastic, forward-looking
     → Vocabulary: Action-oriented, innovative, bold
     → Avoid: Passive language, excessive caution, dullness

  3. COMPETENCE:
     → Traits: Reliable, capable, intelligent, successful, secure
     → Examples: Microsoft, FedEx, Merck, Nordstrom
     → Tone: Confident, authoritative, expert
     → Vocabulary: Precise, data-backed, professional
     → Avoid: Uncertainty, vagueness, casual tone

  4. SOPHISTICATION:
     → Traits: Elegant, premium, charming, prestigious, luxurious
     → Examples: Chanel, Rolex, Ritz-Carlton, Tesla
     → Tone: Refined, aspirational, understated
     → Vocabulary: Elevated, curated, exclusive
     → Avoid: Colloquialisms, discount language, mass-market appeal

  5. ROUGHNESS / RUGGEDNESS:
     → Traits: Tough, outdoorsy, durable, masculine, adventurous
     → Examples: Harley-Davidson, Timberland, Jeep, Dew
     → Tone: Direct, no-nonsense, authentic
     → Vocabulary: Bold, unpolished, straightforward
     → Avoid: Delicate language, excessive decoration, softness

  6. PLAYER / FUN:
     → Traits: Fun, playful, spirited, humorous, entertaining
     → Examples: M&M's, Old Spice, Wendy's, Mailchimp
     → Tone: Lighthearted, witty, irreverent
     → Vocabulary: Casual, creative, surprising
     → Avoid: Formal language, excessive seriousness, corporate speak

YOUR BRAND'S PERSONALITY BLEND:
  → Primary dimension: [Select 1] — 50% of voice DNA
  → Secondary dimension: [Select 1] — 30% of voice DNA
  → Tertiary dimension: [Select 1] — 20% of voice DNA
  → Example: Mailchimp = Player (50%) + Sincerity (30%) + Competence (20%)
  → Example: Stripe = Competence (50%) + Sophistication (30%) + Excitement (20%)

VOICE TONE SPECTRUM:

  FORMAL ←——————— YOUR BRAND ———————→ CASUAL

  → Define your brand's resting position on this spectrum
  → Example: "We're 70% casual, 30% formal"
  → Channel overrides:
     * Website: 70% casual (base position)
     * Social media: 85% casual (more relaxed)
     * White papers: 50% casual (more formal)
     * Press releases: 30% casual (more formal)
     * Customer support: 80% casual (more conversational)

TONE ADAPTATION BY CONTEXT:

  CELEBRATORY (good news, wins, launches):
    → More energy, exclamation points, enthusiastic vocabulary
    → Example: "We're thrilled to announce..." not "We are pleased to inform..."

  EMPATHETIC (support, apologies, difficult topics):
    → Softer tone, personal language, "you" focused
    → Example: "I understand how frustrating this is" not "We acknowledge your inconvenience"

  EDUCATIONAL (guides, tutorials, explainers):
    → Clear, patient, structured, avoiding jargon
    → Example: "Here's how it works" not "The mechanism operates via..."

  URGENT (alerts, deadlines, critical information):
    → Direct, action-oriented, no filler
    → Example: "Action required: Your subscription expires tomorrow"

  AUTHORITY (thought leadership, research, positions):
    → Confident, data-backed, measured confidence
    → Example: "Our research shows 73% of companies..." not "We think maybe..."

Voice Guidelines Document

BRAND VOICE GUIDELINES TEMPLATE
=================================

SECTION 1: WHO WE ARE

  MISSION STATEMENT (voice-influencing):
    "[Company] helps [audience] [achieve result] through [approach]."

  BRAND DESCRIPTION IN 5 WORDS:
    [Word 1], [Word 2], [Word 3], [Word 4], [Word 5]
    Example (Mailchimp): "Colorful, friendly, clever, helpful, human"
    Example (Stripe): "Clean, precise, powerful, modern, developer-first"

  WHAT WE STAND FOR:
    → [Value 1]: How it shows up in our voice
    → [Value 2]: How it shows up in our voice
    → [Value 3]: How it shows up in our voice

SECTION 2: HOW WE SOUND

  WE ARE:
    → [Trait 1]: Description with examples
       Example: "CONVERSATIONAL — We write like we talk. Like we're explaining
       something to a smart friend over coffee."
    → [Trait 2]: Description with examples
    → [Trait 3]: Description with examples
    → [Trait 4]: Description with examples

  WE ARE NOT:
    → [Anti-trait 1]: What to avoid
       Example: "NOT CORPORATE — We don't use words like 'leverage,' 'synergy,'
       'paradigm shift,' or 'best-in-class.'"
    → [Anti-trait 2]: What to avoid
    → [Anti-trait 3]: What to avoid
    → [Anti-trait 4]: What to avoid

  DO / DON'T EXAMPLES:

    DO: "Let's get your store set up in 5 minutes."
    DON'T: "The deployment of your commercial platform shall be effected expeditiously."

    DO: "Here's what changed in version 3.2."
    DON'T: "We are pleased to announce the release of iteration 3.2."

    DO: "You might be wondering how this works."
    DON'T: "It is pertinent to address the operational methodology."

SECTION 3: VOCABULARY RULES

  APPROVED WORDS (use these):
    → [List of brand-specific vocabulary, terminology, product names]
    → [Words that reinforce brand personality]

  FORBIDDEN WORDS (never use):
    → [Corporate jargon: leverage, synergy, ecosystem, holistic, robust]
    → [Hype words: revolutionary, game-changing, world-class, cutting-edge]
    → [Vague words: things, stuff, a lot, very, pretty]
    → [Overused tech words: AI-powered, blockchain-enabled, cloud-native]
    → Example forbidden list: 50–100 words that never appear in brand content

  STYLE RULES:
    → Sentence length: Average 15–20 words (shorter = more casual)
    → Paragraph length: Maximum 4 sentences per paragraph
    → Active voice: Always (not passive) — "We built" not "It was built"
    → Contractions: Yes (we're, don't, can't) — unless formal channel
    → Em dashes: Yes (— like this) — for conversational asides
    → Bullet points: Yes — for scannability
    → Exclamation points: 0–1 per piece (restraint = sophistication)
    → ALL CAPS: Never (shouting = unprofessional)
    → Second person ("you"): Preferred over third person

  PUNCTUATION GUIDELINES:
    → Commas: Use for natural pause in speech
    → Semicolons: Rare (only when sentence structure demands it)
    → Colons: For introducing lists or explanations
    → Em dashes (—): For conversational interruptions, asides, emphasis
    → Parentheses: For side notes, clarifications, asides
    → Ellipsis (...): Rarely (can feel hesitant or manipulative)

SECTION 4: CHANNEL-SPECIFIC ADAPTATIONS

  WEBSITE / LANDING PAGES:
    → Voice: [Base personality]
    → Length: Concise, scannable
    → Tone: Confident but approachable
    → CTA style: Direct, action-oriented

  SOCIAL MEDIA:
    → Voice: [More casual, more personality]
    → Length: Platform-specific (Twitter: short, LinkedIn: thoughtful)
    → Tone: Conversational, engaging, timely
    → CTA style: Engaging question, subtle prompt

  EMAIL MARKETING:
    → Voice: [Personal, one-on-one]
    → Length: Medium, structured
    → Tone: Helpful, personalized
    → CTA style: Clear, benefit-focused

  BLOG / EDITORIAL:
    → Voice: [Expert but accessible]
    → Length: Comprehensive (1,500–3,000 words)
    → Tone: Educational, authoritative, engaging
    → CTA style: Soft, value-adding

  CUSTOMER SUPPORT:
    → Voice: [Most casual, most empathetic]
    → Length: Brief, solution-focused
    → Tone: Patient, understanding, helpful
    → CTA style: Direct, problem-solving

Voice Scoring and Analysis

BRAND VOICE CONSISTENCY SCORING SYSTEM
=========================================

SCORING DIMENSIONS (each scored 0–100):

  1. TONE ALIGNMENT (30% weight):
     → Is the formality level appropriate for channel?
     → Is the emotional register correct (celebratory, empathetic, authoritative)?
     → Does it match the brand personality blend?
     → Score: % of sentences matching expected tone range

  2. VOCABULARY CONSISTENCY (25% weight):
     → Are forbidden words absent?
     → Are approved terms used correctly?
     → Is the vocabulary complexity appropriate (Flesch-Kincaid grade level)?
     → Score: (1 - forbidden_word_count / total_words) × 100

  3. STRUCTURAL PATTERN MATCH (20% weight):
     → Sentence length distribution matches brand pattern?
     → Paragraph length within guidelines?
     → Active vs. passive voice ratio?
     → Use of approved punctuation patterns?
     → Score: % of structural elements matching guidelines

  4. MESSAGING PILLAR ALIGNMENT (15% weight):
     → Does content reinforce brand messaging pillars?
     → Are key value propositions present?
     → Is the brand story/framework referenced appropriately?
     → Score: % of content aligned to messaging pillars

  5. AUDIENCE APPROPRIATENESS (10% weight):
     → Is the content written for the target audience?
     → Technical level appropriate for audience sophistication?
     → Cultural sensitivity checked?
     → Score: Manual review or AI-assisted audience matching

OVERALL VOICE SCORE:
  → Weighted average of 5 dimensions
  → 90–100: On-brand (no changes needed)
  → 75–89: Mostly on-brand (minor suggestions)
  → 60–74: Needs revision (several voice deviations)
  → 40–59: Off-brand (significant rewrite recommended)
  → Below 40: Not on-brand (complete rewrite needed)

VOICE CHECK WORKFLOW:

  AUTOMATED CHECK (before publishing):
    1. Content submitted to voice checker
    2. AI analyzes against brand voice profile
    3. Generates voice score and specific deviation flags
    4. Highlights sentences/phrases off-brand with suggestions
    5. Author reviews and accepts/rejects suggestions
    6. If score < 70, requires manual review before publishing

  MANUAL REVIEW (for high-stakes content):
    1. Brand manager or content lead reviews flagged content
    2. Evaluates context-specific voice decisions
    3. Approves or requests revision
    4. Documents any voice guideline exceptions

  CONTINUOUS LEARNING:
    → Track which voice suggestions are accepted vs. rejected
    → Update voice profile based on approved published content
    → Quarterly voice guideline review and update
    → A/B test voice variations and incorporate learnings

Integration Points

Edge Cases