Section 1
People do not process a leader's story neutrally
Three observable behaviours shape how internal messages land, and you can watch all three inside your own company this quarter. The first is gap-filling under threat. When the stakes are personal, ambiguity does not read as ambiguity. It reads as bad news being withheld, because employees have usually seen bad news withheld before. The second is first-framing stickiness. Whichever explanation reaches people first becomes the reference version, and later corrections attach themselves to it as amendments rather than replacing it. The third is status arithmetic. Every reorganisation announcement is scanned for who gained scope and who lost it, regardless of what the announcement says it is about. None of this is irrationality. It is reasonable inference from limited information by people with real exposure to the outcome. Treating it as a communications failure to be trained out of them is where most leaders go wrong, and [Why Storytelling Is a Critical Leadership Skill for Founders](/blog/why-storytelling-is-a-critical-leadership-skill-for-founders) makes the case for treating it as a design constraint instead.
Section 2
What follows for how you sequence a message
If gaps get filled with the worst reading, the sequencing rule writes itself: name the hard part first, in plain words, before the reasoning. People cannot hear the reasoning while they are still working out whether they are affected. That means saying whether jobs are at risk, whether the roadmap is being cut, and when the uncertainty will be resolved, even when the honest answer is that you do not know yet and will say more on a stated date. An unresolved question with a date attached is tolerable. An unresolved question with no date gets resolved privately by the audience, and once it is resolved privately you are arguing against a version of the story you never heard. Reassurance offered before the hard part is stated is not received as reassurance. It is received as evidence that the hard part exists.
Section 3
Matching the message to the reaction
The table below pairs the common internal announcements with the inference the room will make by default, and the specific sentence that closes the gap. Use it before the meeting rather than in the follow-up email, which is where most of these sentences currently appear.
Section 4
Repetition, and why it feels excessive
By the time a leader is bored of a message, most of the organisation has heard it roughly twice, in different states of attention, and a portion was on leave. The gap between how often you have said something and how often it has been received is the single most reliable source of surprise in internal communication. The practical response is to plan repetition as a sequence with different framings rather than the same words repeated. The all-hands establishes the decision and the cost. The team meeting a week later works through what it changes in that team's actual week. The written version, which people forward and search later, carries the detail neither meeting had room for. New joiners then receive it as history rather than news, which is a different message with different requirements, taken up in [Using Storytelling in Employee Onboarding and Training](/blog/using-storytelling-in-employee-onboarding-and-training).
Section 5
The manipulation line
These mechanisms work whether or not what you are saying is true, and that is the honest problem with the topic. A leader who understands how gaps get filled can also use narrative to keep people calm while a decision is made against their interests, and it will work for a while. It fails on the second cycle. Employees compare the last story to what actually happened, and a mismatch resets the credibility of everything else you have said, including the parts that were accurate. The safeguard is boring: only make claims about the future that you would put in writing with a date. Automation announcements are where this line gets crossed most often, since the reassuring version and the accurate version usually differ, and [How AI Automation Changes Business Models, Margins, and Speed](/blog/how-ai-automation-changes-business-models-margins-and-speed) lays out what is actually at stake.