messaging-and-tone-matrix

$npx mdskill add Significant-Gravitas/skills-catalog/messaging-and-tone-matrix

Use this once a voice already exists and the question has moved from *how do we sound* to *what do we say, to whom, and where*.

SKILL.md

.github/skills/messaging-and-tone-matrixView on GitHub ↗
---
name: "messaging-and-tone-matrix"
description: "Build the layer above the voice guide: the positioning line, the three proof points you repeat, the claims you will never make, the language each persona gets, and where the tone sits for sales, support, marketing and social."
triggers: ["messaging framework", "proof points", "tone by channel", "claims we cannot make", "persona messaging", "words we never use"]
version: "1"
---

# Messaging and tone matrix

Use this once a voice already exists and the question has moved from *how do we
sound* to *what do we say, to whom, and where*.

This skill does not capture the voice. Sentence length, contractions,
headings, and punctuation habits belong in a voice guide: if one exists, read
it first; if none does, pull the three or four voice rules you can see in
their own writing, note them at the top, and say the full guide is a separate
job. If a positioning read already exists — from an ideal customer profile
(ICP) exercise or a GTM strategist — start from it rather than writing a
second one that disagrees.

## The positioning line

One sentence, in this shape:

> For [who], [product] is a [category] that [benefit]. Unlike [alternative], we
> [proof].

Every blank is filled from something they told you or something published. A
blank you cannot fill stays blank and becomes the first question back.

## The three proof points

Three things the company repeats, each with a quoted example or a named source
under it. A proof point you cannot show an example for gets cut rather than
softened. Three is the cap: a list of nine is a list nobody uses.

## The claims we never make

Write them down explicitly. This is the most useful page in the whole document
and the one most teams never write: the numbers they have not measured, the
comparisons legal will not sign, the guarantees the product cannot keep, the
competitor claims they will not make. Each with one line on why.

## Per-persona variants

When they serve more than one buyer, write a variant per persona: same
positioning, own language, own proof points, own call to action (CTA). Do not
write a variant for a persona nobody can name — ask first.

## Two word lists

Ten to fifteen words to use, ten to fifteen to avoid. The avoid list is the
corporate words they already never say, plus the hype words and the filler
openers every draft gets cut against. Both lists come from their own writing,
not from a generic list of bad words.

## The tone-by-context matrix

One line per context, each with a quoted example, saying where the voice sits
on the formal-to-casual spectrum:

- **Sales** — what a first touch and a follow-up sound like.
- **Support** — what an apology and a refusal sound like.
- **Marketing** — what a page and a campaign email sound like.
- **Social** — what a post and a reply to criticism sound like.

A context with no example yet gets marked unwritten rather than guessed.

## How to keep it

Show the whole thing in chat, take one round of edits, then save it. Rewrite it
rather than appending to it whenever their edits contradict the same rule
twice. Keep it short enough that a new writer reads all of it.

## What not to do

A claim, customer logo, or statistic you found on someone's website is not
proof you can reuse; only what the user confirms goes in the matrix. Never
research a company they did not name. Never write a proof point the product
cannot demonstrate.

More from Significant-Gravitas/skills-catalog

SkillDescription
account-health-and-qbrsUse when a support account wobbles or a quarterly business review looms: read the health signals, run the success plan, and prep the review from evidence.
accounts-receivable-follow-upReview open receivables and draft factual, staged payment follow-ups without inventing status or contacting a customer.
ad-copy-variantsWrite ad variants that each test one idea, within platform limits and supported claims.
alex-getting-startedUse on the first conversation with Alex, or whenever their memory has no product preferences yet: learn what the user is building and who for, where specs, roadmap and numbers live, who decides dates and scope, and get them to a first real product deliverable.
alliance-co-commercializationUse when a strategic alliance needs joint selling governance: operating model, joint targeting, steering prep, and milestone accountability.
anika-getting-startedUse on the first conversation with Anika, or whenever their memory has no partnership preferences yet: learn which partners and alliances the user owns, what motion they run, and get one real partner read on screen in the same session.
assure-partner-led-deliveryUse when partners deliver client work in your name: own the in-flight book, run the weekly delivery review, and rescue engagements before clients feel it.
automate-finance-reportingUse to connect a number source, map an export into the finance ledger, or QA a sheet: the column mapping, the dedupe key, the load summary, and the checks that must pass before a read ships.
billing-refunds-and-exceptionsUse when money is on the table: verify the charge, check the policy, and stage a refund or exception draft that stops at the owner's yes.
board-and-investor-metrics-briefPrepare a concise board or investor metrics brief with definitions, sources, comparisons, drivers, risks, and decisions needed.