Message matrix: meaning and application

Address several target groups and you soon have texts that are each correct but drift apart. How do you record, per target group, what you say, how you support it and what you avoid? You learn where the message matrix comes from, how to fill in supporting messages, fixed evidence and a pitfall sentence, and its limits.

A message matrix is an overview that records, for each target group, the core message, the supporting messages that make it concrete, the evidence that always goes with it and the wording you avoid. The message map is the elaboration of that overview for a single target group.

The tool comes from communication practice and is related to the message house and to risk communication's message map. At Neurofactor, every cell refers to a researched target group insight. The matrix organises messages; only testing shows whether they work.

A message matrix is a practical tool from the communication profession, not a scientific concept with one original publication. Two related forms can be traced.

Message map in risk communication. Vincent Covello, by his own account, developed the message map in the early 1990s for high-stress, emotionally charged situations; after the 2001 anthrax attacks in the United States it was first adopted in public health (Covello, 2006). A message map answers one question from a stakeholder with three key messages and supporting facts. A report for the US Environmental Protection Agency (EPA) describes seven steps, ending with testing and practising the messages, and notes that not every stakeholder group needs its own map (Covello et al., 2007).

Message house. A message house groups several messages under one 'roof'. According to CommunicatieRijk, the Dutch central government's communication platform, it provides an overarching narrative when different messages are used for different target audiences (CommunicatieRijk, 2023). A definite origin was not found in the sources consulted.

Neurofactor's interpretation. Neurofactor uses 'message matrix' for the overview of all target groups side by side and 'message map' for the elaboration per target group. Specific to this working definition: every cell refers to the insight it rests on, and the pitfall sentence is a fixed column. It is an editorial agreement, not a validated model.

One row per target group, one column per component:

The core message, the starting point of each row, will get its own article. The matrix records what you say; how you say it belongs to tone of voice. A message map elaborates one row, with wording per channel, sources, evidence owner and last check date.

A supporting message is a statement that makes the core message concrete, understandable or credible for a specific target group. It answers the follow-up question: how does that work, and why should I believe it?

Covello's message map puts three supporting facts under each key message, and the EPA report describes a template of three key messages totalling 27 words, spoken in nine seconds (Covello et al., 2007). These are conventions for spokespeople in crisis situations. No fixed number is substantiated for marketing or HR communication; two supporting messages per group is a common choice that keeps the overview readable.

When choosing, check three things:

• Traceable. Each rests on a target group insight, such as a pain point or objection.
• Own substantiation. A supporting message is itself a claim. See also What does evidence mean in a target group profile?
• Different per target group. If the supporting messages are identical everywhere, the matrix may not capture real differences.

The fixed piece of evidence always accompanies the message for a target group, whatever the channel or text length; hence the working name 'always show'. In a short advert much is dropped, but this evidence stays or is directly accessible.

Why one fixed element? If evidence varies by channel, someone may never see the evidence for their main doubt. This is an editorial judgement, not a researched rule.

Criteria (editorial advice):

• It addresses the group's main doubt or objection.
• It is verifiable and current, with an owner and a check date.
• It can be shown or reached in every channel.
• It promises no more than it shows.

What makes evidence credible is a separate topic; the matrix only records what always goes along.

A pitfall sentence is a concrete sentence you deliberately avoid for a target group, with the reason stated. It works as a counter-check: if a draft resembles it, you rewrite the draft. Unlike a word list, it records an assumption about how the group may read a phrasing.

Research shows wording can matter, though not one-to-one for every text:

• Framing. Levin, Schneider and Gaeth distinguish three types of framing effects, including attribute framing: the evaluation of the same attribute can differ depending on whether it is described positively or negatively (Levin et al., 1998).
• Negations. In two experiments with students, negating an attribute without a clear opposite ('not romantic') was processed differently from one with a clear opposite ('not warm'): the first kind activated associations with the negated concept. The authors advise describing what something is (Mayo et al., 2004).
• Repeated warnings. Older adults who repeatedly saw a consumer claim marked false more often remembered it as true three days later than after one warning; younger adults did not (Skurnik et al., 2005).
• No general ban. A review concludes that backfire effects of corrections are not a robust phenomenon and rarely, if ever, occur with a clear and salient correction (Swire-Thompson et al., 2020).

Covello advises avoiding unnecessary absolutes and words such as 'no', 'not' and 'never' (Covello, 2006). The rule of thumb that one negative needs three positives appears in the EPA report without a reference (Covello et al., 2007); this article therefore does not adopt it as a norm.

A pitfall sentence is thus a hypothesis about wording. You can test the assumption behind it in advance, as described in From target group insights to testable messages.

Three related articles cover other levels: the associative value proposition (what you promise and why), claim, evidence, clarification and language anchors (how to write out one promise) and the evidence anchor (the evidence itself).

1. Start with target groups and choice context. Take them from the target group profile or research. See also From target group profile to communication strategy.
2. Formulate the core message per target group. One sentence the group should remember.
3. Choose supporting messages. Use insights that raise the most questions for this group.
4. Assign the fixed evidence. Check that it exists, is current and works in every channel.
5. Write the pitfall sentence. Note a concrete sentence and the reason.
6. Record origin and status. Note each cell's source and whether it is tested.
7. Test and update. Pretest messages with the target group; revise the matrix when insights or the offer change.

This example is fictional. The table is a working hypothesis, not a research outcome.

Situation. A university of applied sciences offers a part-time bachelor's in business administration to three groups: early-career workers without a degree, returners after a career break, and managers who want to underpin their experience.

Choice question. Which message, evidence and wording suit each group, so all channels tell the same story?

Available information. Conversations with student advisers, questions from open days and a first target group profile.

Suitable approach. The team fills in a message matrix, noting the insight behind each cell.

Possible interpretation. Each group seems to have a different doubt: time, fit with previous experience, applicability. Whether the messages remove them is unknown.

Next step. The team pretests the messages per group with predefined outcome measures and checks that the fixed evidence is available in every channel.

• Treating the matrix as evidence. A matrix shows what you want to say, not that it works.
• Using numbers as a norm. Three key messages or 27 words are risk communication conventions, unsubstantiated elsewhere.
• The same evidence everywhere. Doubts differ per group, so the fixed evidence usually does too.
• Pitfall sentences as a word ban. A negation is not harmful by definition; effects depend on context and wording.
• Stretching findings. The experiments involved specific tasks and groups; they do not predict one reader's response.
• Unsourced or outdated cells. A message without a source is an assumption; when the offer changes, messages and evidence age with it.

A message matrix brings together, per target group, what you say, how you support it and what you avoid. Neurofactor adds traceable insights and a pitfall sentence as a counter-check. Treat numbers as a common choice, not a norm, and every cell as a hypothesis until tested.