After nearly every serious institutional failure, it emerges that several people knew something was wrong. This is so consistent that it should be treated as a design fact rather than as a discovery.
The question worth asking is not why nobody noticed. It is why what was noticed did not travel.
Signals arrive as anomalies
Nothing announces itself as a warning. What arrives is an oddity with a plausible explanation already attached.
A near miss that produced no consequence, so there is nothing to report. A number that is off and gets reconciled. A customer complaint resolved to their satisfaction. An experienced person resigning with a reason given that nobody examines. A supplier requesting different payment terms.
Each is individually unremarkable and each explanation is offered in good faith by somebody who believes it. What makes them signals is not any one of them but the fact that four arrived in the same quarter, and no one person saw all four.
The near miss is the most wasted information
An event that nearly caused harm and did not is the cheapest possible instruction, and most organizations discard it entirely.
The reason is structural. A near miss produces no damage, so there is no incident to investigate, no insurer involved, and no obligation to record anything. Reporting it therefore requires somebody to volunteer that they nearly caused a problem, which is a purely costly act.
Industries where the consequences are lethal solved this long ago by making near-miss reporting routine, non-punitive, and expected. Most ordinary organizations have not, and the near misses that would have told them what was coming are known individually to the people involved and to nobody else.
The intervention is small: a standing, low-effort way to record something that nearly went wrong, with an explicit rule that recording it is never held against the person recording it, and evidence that the rule holds. Without the last part the first two are decorative.
Explaining is the mechanism of suppression
Signals are rarely ignored. They are explained, which feels like the opposite of ignoring and produces the same result.
The variance was timing. The complaint was an unusual customer. The resignation was personal. The near miss was a one-off caused by conditions that will not recur.
Every one of these may be true, and each is offered by somebody with a mild interest in the matter being minor: the person responsible, their supervisor, or anyone who would otherwise have to escalate.
The useful discipline is to notice when an explanation is doing a lot of work, and to ask one further question. Has this happened before? That question is answerable, it costs nothing, and it converts a single anomaly into a pattern or dismisses it properly.
Nobody holds all the pieces
The characteristic feature of a missed crisis is that the information existed and was distributed.
Finance saw the receivable slipping. Operations saw the quality complaints. HR saw the resignations. The site supervisor saw the near miss. Each addressed their own item competently within their own function, and no forum existed in which the four appeared together.
This is why the correction is a meeting rather than a report. A short standing item where each function names anything unusual, without a requirement that it be significant, produces the assembly that no individual report can. The value is not in any single contribution; it is in two of them turning out to be about the same underlying thing.
What early recognition buys
Not avoidance. Most of what is coming arrives regardless of when it is spotted.
What it buys is options, and options decay fast. A financial problem identified in month two can be addressed by a range of ordinary measures. The same problem in month nine has narrowed to two unpleasant choices, and in month fourteen to one. Nothing about the underlying situation changed; the available responses expired.
It also buys the ability to disclose on your own terms, which is worth a great deal and is the first thing lost. An organization that raises something with its board, its funder, or its regulator before being asked is in an entirely different position from one that is discovered, and the difference is not the substance.
Which makes early recognition less a matter of vigilance than of arithmetic. Every week of delay is a reduction in what you can still do, and the reduction is invisible until you go looking for a response and find it gone.
Edited by Patrick J. Wolf, PhD