Build a voice brief that changes actual writing decisions

Choose the reader situation and permitted material

A useful brief helps a writer decide how to explain a reporting change, respond to a problem or correct a number. Start with the intended reader and their practical question. Select a few own or expressly permitted pieces that demonstrate the work you want to maintain. Keep their actual context and version; a public profile is not a license to copy private experience or claim another person’s voice as your identity.

Mailchimp distinguishes relatively stable voice from tone that changes with context. That is its publisher-specific model, not a universal personality taxonomy. We adapt the distinction into a creator writing exercise. No real author was interviewed, no audience preference was measured and no software inferred a person’s identity from their prose.

Write rules that imply an observable edit

Replace “authentic, bold and friendly” with decisions. Our constructed brief says: name the concrete change; explain one unfamiliar term when needed; retain numbers with their units and limits; offer the next useful step without pressure. Each rule has an exception: technical terms may be necessary if defined, and a serious correction may need direct restrained language rather than a joke.

Separate style preference from factual review. “Use a warmer opening” cannot authorize an invented customer story, a stronger causal verb or an absent result. Keep source checks and approval responsibilities in their own records. A brief can help reviewers discuss an edit; it cannot grant access to an account or establish that the final publication was approved.

Inspect three complete tone changes

Constructed announcement: before, “Our revolutionary update guarantees effortless reporting.” After, “The worksheet now includes a source note beside each field. Open the example to see which questions the supplied data can answer.” The edit names a change and a usable destination without inventing effort saved. The destination must still be checked in real use.

Constructed explanation: before, “Obviously, anyone can see that the total is not the result.” After, “This total counts requests. It does not identify how many people completed the task. Keep those measures separate when you describe the report.” Constructed correction: before, “A tiny tweak: everything is fixed.” After, “Correction: the earlier caption used the wrong period. The revised caption names the supplied window; the reader outcome remains unmeasured.” These are original fictional drafts, not real incidents.

Resolve three voice-review decisions

V1 — Constructed unsupported confidence: an editor asks for “more authority” and the rewrite claims independent validation. Decision: keep the actual review scope and improve clarity instead. Outcome: a clearer supported claim, no new reviewer or credential.

V2 — Constructed forced humor: the normal brief allows light humor, but the reader is reporting a failed download. Decision: acknowledge the concrete issue and give the relevant recovery step without assuming how the person feels. Outcome: situational tone, not a diagnosis or a promise that the problem is resolved.

V3 — Constructed imitation: a draft copies a recognizable creator’s phrases and first-person achievements. Decision: write from your own permitted material and supported contribution. Outcome: an owned explanation, not a human-origin certificate or a borrowed career history.

Review a whole draft and a genuine exception

Use the rules on a complete current draft, not only a headline. Mark the exact proposed edit and its reason; then reread factual scope, source access and the next step. If two reviewers disagree, preserve their actual reasons and decide which rule or exception applies. An automated style score does not settle meaning or create an independent reviewer.

Keep examples in each intended language. A literal slogan can sound unnatural or lose a qualification in German. Review the whole counterpart and note deliberate differences. Do not assume that identical punctuation, sentence length or emoji use creates the same relationship with every reader. The original translation and quote-fidelity guides cover those separate checks.

Maintain a small usable brief

Record reader situation, permitted sample/version, rule, reason, exception, complete before/after wording, protected facts and unresolved review. The download contains the full examples and a blank private brief. The checker verifies that three declared scenarios retain their factual boundary; it does not judge tone, identity, originality, rights, trust or engagement for new text.

The primary style-guide page was read on 1 November 2026. All writing examples and V-cases are constructed. No author interview, profile change, collaboration, message, native post or reader experiment was performed. Revise the brief when genuine use reveals a problem; do not label it reader-validated before that evidence exists.

1. Select permitted samples and their actual context.
2. Write edit decisions with reasons and exceptions.
3. Draft announcement, explanation and correction examples.
4. Check protected facts across languages and tone changes.
5. Record actual review and revise the brief when warranted.