Content tone matrix

Description

A grid mapping user situations and emotional states to the tone a product should take, audited against Gricean maxims so wording implies only what it should.

A content tone matrix crosses the situations a product speaks in (onboarding, success, error, payment failure, account lock, bereavement) with the emotional state the user is likely in and specifies the tone for each cell along defined dimensions, typically formal to casual, serious to playful, respectful to irreverent, enthusiastic to matter-of-fact. Its rigour comes from two sources. First, the cells are derived from evidence about user state (support tickets, interviews, diary entries) rather than brand aspiration. Second, each sample string is audited pragmatically: does it say as much as needed and no more (Quantity), is it true (Quality), relevant (Relation) and clear (Manner)? Because users read implicatures, not just literal meaning, an error message that is jaunty or vague can imply blame or triviality. The matrix turns tone from a stylistic preference into a testable specification.

Attributes

Loss type
Active
Stage
Design
Data
Mixed
Effort
Days

Steps

  1. Inventory every message context in the product (aim for full coverage, typically 30 to 120 message types) and group them into 8 to 15 situation types such as success, recoverable error, blocking error, money movement and irreversible action.
  2. For each situation, estimate the user's likely affective state and stakes using evidence: support tickets, session recordings and at least 5 to 8 interviews with users who recently experienced that situation.
  3. Define 3 to 5 tone dimensions with anchored endpoints (for example NN/g's funny-serious, formal-casual, respectful-irreverent, enthusiastic-matter-of-fact) and set a target position on each for every situation row.
  4. Write 2 to 3 exemplar strings per cell and run a Gricean maxim audit on each, flagging any string that over- or under-informs, asserts something not guaranteed, is irrelevant to the user's goal, or is ambiguous.
  5. Test the highest-stakes cells (errors, money, health, security) with 5 to 10 users using a comprehension and implicature probe: 'What is this message telling you? Whose fault is it? What will you do next?'
  6. Revise cells where more than 1 in 5 participants draw an unintended implicature (for example self-blame or false reassurance), then publish the matrix in the design system with do and do-not examples.
  7. Re-audit quarterly or whenever new message types ship, sampling live strings against the matrix.

Sources

  • Development of Language (Gleason)
  • Conversational Design (Hall)
  • Methods Atlas — Gricean audit (Gopinath)
  • The Inquiry Series, Ch 18 — Pragmatics — What Language Actually Does

Also known as

  • Voice and tone matrix
  • Tone decision matrix
  • Tone map
  • Context-by-tone grid

Categories

  • Content
  • Linguistics, Semiotics & Discourse