Pattern briefs
Short reads, one pattern each. Every one names the pattern and explains the mechanism. If it sounds like your week, the next step is a week of investigation, not another article.
Almost every organisation with ten or more staff has an Internal Committee on paper. Very few have tested whether it can run an inquiry. The names are listed, the policy is filed, and the first case the committee handles badly is the one that reaches a lawyer. The pattern is not negligence. It is that compliance is measured by what exists rather than by what works, so a constituted committee reads as done. What is missing is the rehearsal: the ninety-day clock, the evidence, the interim protection, the written record that would hold if it were examined. None of that is tested until it matters, and by then it is too late to be good at it.
Somewhere in the last year, a written process stopped being followed and a workaround took its place. It is faster, it survives the busy weeks, and it is undocumented. Nobody decided this. It accumulated. And now the organisation runs on a process that exists only in the heads of the people who invented it, which means the rework, the defects and the risk all live in the gap between the document and the floor. The written process is current and wrong. The real one works and is invisible. The moment someone leaves, the real one leaves with them, and the organisation discovers the gap only when it is expensive.
A metric that is reported by the people whose work it describes, and rewarded by the people who want it green, will eventually be managed rather than measured. It does not take dishonesty. It takes a system that has learned, reasonably, that a green number is safer than an honest one. So the exceptions get handled off the record, the definition drifts, and the dashboard slowly stops describing the floor. By the time the two disagree, the disagreement has usually been there for a year, and the people closest to it can no longer see it, because it is the water they swim in.
When the same person leaves in the same month for the same reason, the cause is not the person. It was set at the interview. What the candidate was told about the role, the growth and the team is rarely what they meet on day one, and the gap between the two is what they leave over. The interview itself usually tests what is easy to score rather than the thing the role actually fails on, so the panel is confident and wrong. And the first ninety days are handed over and forgotten until the review, by which time the decision to leave has already been made.
Training is one of the few large budgets that is almost never audited for effect. Attendance is counted, feedback is collected, and the question of whether anything changed on the floor is left unasked. So a session can be full, well-rated and useless, and nobody can tell the difference between that and one that landed. Some training should not exist. Some should be repeated differently. Some works and is never measured, so the organisation cannot tell which is which, and the budget keeps buying the same thing.
The exit interview is a form filled in by someone who has already decided to leave, and it is almost always polite. The real pattern is not in what they said. It is in the record around the departure: who left, after how long, from which team, and what changed in the months before they went. Attrition figures give you the rate and hide the mechanism. The same seat, refilled every eighteen months, is a pattern rather than bad luck, and it is visible only when the exits are read as a set rather than as individual stories.
The full essay, and what it looks like on a real floor.
If one of these sounds like your last month, the next step is a week of investigation, not another article.
If the pattern is real, you will know within a week. If it is not, we say so in writing.
Request a Proposal