Verbal Identity

When to use this

Use when the brand's writing does not sound like anyone in particular, when different teams produce work that could not have come from the same organization, or when existing voice guidance is a list of adjectives nobody knows how to apply.

Use when the organization is scaling its writing beyond the people who intuitively know how the brand sounds. That intuition does not survive hiring.

Do not use it to write a single campaign. This produces the system a writer works inside.

Inputs

The position, promise, distinctive mechanism, and proof inventory from strategy-brief.md. The name and its constraints from naming-report.md where naming was in scope. Existing writing across at least product UI, support, sales, and marketing — including the writing people are embarrassed by, which is more diagnostic than the writing they are proud of.

If you have no existing writing to read, ask for some before defining a voice. A verbal system built without seeing how the organization currently writes describes an aspiration rather than a change, and nobody will be able to tell what it asks them to do differently.

Mission

Make the brand sound like a distinct entity while remaining clear, credible, and appropriate to what the reader is actually going through.

Voice is not a mood. It is a consistent set of decisions about what this brand notices, admits, refuses, explains, and asks for. Specificity creates voice. A point of view creates recognition. Casual language creates neither.

Method

1. Separate the five layers. Voice (stable character), tone (contextual variation), message (what is said), proof (why it is believed), action (what is asked). Most broken verbal systems have collapsed two of these.
2. Define character through worldview, relationship to the reader, level of confidence, and where humor stops. Derive each from the strategic position, not from writing preferences.
3. Set voice tensions. Position the brand on real dimensions — direct/considered, plain/technical, warm/exacting — and say why. A tension has two credible ends; if one end is obviously wrong, it is not a dimension, it is a platitude.
4. Build the tone matrix. For each context, name the reader's condition, what shifts, and what must remain stable. Tone that varies without a stable core is not a voice.
5. Build the messaging hierarchy: master idea, promise, supporting messages, proof, detail, action. Each level must survive being read alone.
6. Write the proof language rules. How this brand communicates evidence without overclaiming, and how it handles the cases where the evidence is thin. This is where most brand voices lie.
7. Define language behavior: vocabulary the brand uses and avoids, syntax, punctuation, jargon policy, how it handles uncertainty, apology, refusal, and bad news.
8. Set accessibility standards for language: reading level, sentence length, plain-language requirements, and how the brand writes for translation and for screen readers.
9. Write the same message across every context — homepage, product UI, error state, support reply, sales email, executive communication, social, campaign. This is the deliverable that makes the system usable.
10. Write anti-examples. Copy that is plausible, competent, and wrong for this brand, with the reason. Anti-examples teach faster than rules.

What must be established

Rules that produce decisions. "Be human" is not a rule. "When we do not know, we say we do not know and give the date we will" is a rule — it can be followed, and a reviewer can tell whether it was.

Every constraint should trace to the strategic position. A vocabulary ban with no strategic reason is a preference imposed on people who did not consent to it.

Strong versus merely styled

The system works when a new writer produces something that feels like the brand without copying a phrase from the guidelines.

Failure modes

• Confusing casual language with personality. Every brand went casual; it is no longer a position.
• Forcing humor into contexts where the reader is frustrated, worried, or losing money.
• Building giant banned-word lists. They are unenforceable and they read as contempt for writers.
• Making every channel sound identical, which is consistency without judgment.
• Writing guidance only for marketing, leaving product UI and support — where most words are actually read — ungoverned.
• Optimizing distinctiveness at the cost of clarity. A reader who cannot understand you has not been charmed.
• Producing rules with no examples, which means each team invents its own interpretation.

Never claim without evidence

Never write proof language that asserts results, rankings, certifications, or customer outcomes the organization has not established. Where the strategy flagged proof as still-to-be-built, the verbal system must say what the brand may claim in the meantime.

Never invent a customer testimonial, quotation, or use case as an example, even illustratively. Use the brand's real material or clearly label the example as constructed.

Output

Produce verbal-system.md using the structure in references/verbal-system.md.

Verification

• Blind test. Strip brand names from your copy and from three credible competitors'. Ask someone to identify yours and say what gave it away. "It's the friendly one" is a failure.
• Clarity test, run separately. Comprehension is tested independently of personality, because a voice that costs understanding is not paying for itself.
• Pressure test. Write an outage notice, a price increase, and a refusal. Weak verbal systems have nothing to say here and default to corporate neutral.
• Transfer test. Give the system to a writer who has never worked on the brand. Their draft reveals which rules are real and which are decorative.
• Counterfactual pass. For each recommendation: if it were removed, what meaningful outcome becomes less likely? No credible answer means cut it, downgrade it, or reframe it.
• Fresh-eyes pass. Read the output as someone who must act on it and was not in the process. Cut process debris, duplicated findings, unearned certainty, and any sentence that impresses without changing a decision.

Complete when

Another skilled writer can create new work that preserves character, clarity, credibility, and context sensitivity — without copying phrases from the guidelines.

Handoff

Identity System needs the verbal register to match visual behavior; a precise voice with a loose visual system reads as two brands. Brand Worldbuilding turns messaging patterns into recurring narrative behavior. Brand Governance classifies which verbal rules are core and which are flexible — bring an opinion, since writers will otherwise be governed on everything equally. Brand Critic will test recognition with the name removed, using the blind test above.