The number that ended the firm was in a model nobody had finished
Senior attention on the volume, and the exposure sitting in the part of the work nobody owned.
The hidden performance cost paid when leadership focus is fragmented across too many simultaneous priorities. The calendar, not the strategy deck, tells the truth.
Nothing was concealed and nothing was ignored. The number that ended the firm had simply stopped being read by anyone who could act on it.
What happens
A trading firm cuts its risk desk. On his way out, the head of that desk hands a junior analyst a drive with an unfinished model on it.
The analyst finishes it overnight. The position it describes is larger than the firm is worth, and the limits it breaches have been breached for days.
The reading
Nobody hid anything. The exposure was in the firm's own numbers the whole time, and the model that would have read it was one person's side project in a function that had just been reduced.
Watch how far the answer has to travel once it exists. It goes analyst to desk head to division head to the executive floor, in the middle of the night, because there was no standing route for a risk finding to arrive on its own.
The limits had been exceeded repeatedly and each time became the new normal. That is the tell: the organization was not ignoring an alarm, it had stopped hearing that number as an alarm at all.
Side by side
| Where to look | What the firm did | What attention discipline looks like |
|---|---|---|
| Who owns the worst number | A risk desk being reduced, and one person's unfinished side project | A named owner senior enough to stop the business, with the finding read on a schedule |
| How a breach reads | Exceeded repeatedly, and each time becoming the new normal | The second breach is treated as more serious than the first, not less |
| How the finding travels | Analyst to desk head to division head to the executive floor, overnight, because no route existed | A standing route that does not require somebody to agree it is worth passing on |
What would have changed it
Attention discipline here is not more oversight. It is that the single number capable of ending the firm has an owner senior enough to act on it, and a place it is read out loud on a schedule.
The firm had that function and cut it, which is a decision about attention rather than about headcount, and nobody appears to have framed it that way in the room where it was made.
Where you have seen this
The person who understood your worst exposure left in the last reorganization. The model is still running and nobody senior has read its output since.
What to do about it
A reading that ends in understanding is entertainment. This is filed under Training, so it ends in things a team can actually do.
Name the single number that could end your business, and name the person who reads it. If those are two different answers, or if either one is vacant, that is the finding.
Look at what your last reorganization removed rather than what it added. A function cut is a decision about where attention will no longer go, and it is rarely argued that way in the room.
When a threshold is breached twice, escalate it harder the second time rather than softer. Drift is what a repeated exception feels like from the inside.
What the reading rests on
Disaster incubation, and the normalization of a repeated exception
Two literatures meet here. The incubation account holds that large failures accumulate over long periods through unnoticed events that are at odds with the organization's own beliefs about its safety, so the surprise at the end follows a chain of warnings the organization held and did not read. The normalization account, drawn from a close reading of a launch decision, describes how each accepted anomaly widens what counts as acceptable, until a signal that would once have stopped work is filed as expected variation.
- Turner, B. A. (1976). The organizational and interorganizational development of disasters. Administrative Science Quarterly, 21(3), 378-397. semanticscholar.org
- Vaughan, D. (1996). The Challenger launch decision: Risky technology, culture, and deviance at NASA. University of Chicago Press. bibliovault.org
What this is not
These are readings, not evidence. A film is written to be satisfying rather than to be true, so no screening is a finding about any real organization and none is offered as one. What a screening does is name a mechanism that organizational research has actually studied, cite that research so you can go and check it, and then show the mechanism operating in footage you have already seen. The citation stands behind the mechanism; it never stands behind the film, and pairing the two is our reading rather than a result. Argue with the reading. That is the point of using material you already hold.
You have seen it here
Now read your own organization
A screening is practice. Nothing here measures your organization and nothing here commits it. The diagnostics do the measuring, on your conditions rather than somebody else's.