The Judgment Gap what closes it, and what that costs › Reading four quarters of a gap as it widens 12 steps, no labs
12 — part 4, what closes it, and what that costsexhibit

Reading four quarters of a gap as it widens

Here is four quarters of one organization, and your job is to say what was happening and what you would have funded.

Four quarters of one organization. Read it before the commentary and answer two things: what was happening, and what you would have funded at the end of the second quarter.

One platform organization, four quartersconstructed for this volume
work produced per quarterup 64% across the four
work examined per quarterflat, within noise
produced-to-examined2.1, then 2.8, then 3.6, then 4.4
throughput, as reported upwardthe same rise, reported every quarter, correctly
people doing the examiningunchanged at 4
reworknot computed
what leadership was toldthat the team had had its best year
the fourth-quarter incident11 days to attribute, no single change found

Constructed for this volume. The numbers describe nothing real and are here to be reasoned about, not cited.

both series, on one axis, moving together first second third fourth throughput, as reported produced-to-examined, which nobody reported the same four quarters a reader who sees only the pink series has half of the exercise. The point is that they rise together.
Both series on one axis, moving together. A reader who sees only the ratio has half of the exercise.
What moves: four pairs of bars appear, reported throughput and the produced-to-examined ratio, both rising across the four quarters.

The reading

what you can sayand what follows
the examining capacity never movedwhich it could not: the same people had the same hours. This is the pinned quantity
the ratio roughly doubledso by the fourth quarter most of what shipped had not been looked at by anybody who could have caught a problem
everything reported upward improvedcorrectly, every quarter. Nobody misled anybody
the incident could not be attributedwhich is the signature from step six, and is what the accumulation looks like when it surfaces

What this volume would have funded, and where it could be wrong

At the end of the second quarter, with the ratio already visibly moving: structure the highest-volume output so it can be checked rather than read — the second row of step ten. That is engineering work with a clear payback and it does not need anybody's permission beyond yours.

Not more reviewers. The sum from step seven says the capacity needed had roughly doubled over the period, the people who could do it were the four already doing it, and nothing would have shown whether adding a fifth helped.

Two places this reading could be wrong, stated because a volume that publishes only its confident conclusion is not worth marking against. The examined count may have been flat because examining got faster rather than because capacity was pinned — that would be good news and the exhibit cannot distinguish it. And a single incident is not a pattern; reading the fourth quarter's incident as the bill for the first three is the most likely error on this page.

→ Compute your ratio again, and put the two readings side by side. If you took the first one in step three, you now have a direction rather than a level — and a direction is what this volume is actually about. Then name which of the four moves in step ten you could start without asking anybody.

Every term this page uses

produced-to-examined
How much work your team produced, divided by how much of it somebody actually looked at.
examined
Looked at by somebody with enough context to have caught a problem. Not the same as having passed through a review step.
remainder
Work that was produced and not examined. It does not go anywhere — it merges into what everything after it is built on.
judgment
Deciding whether a piece of work is good, which needs knowing the domain well enough to recognize a wrong answer.
inspection
Checking finished work at the end, rather than changing how the work is made.
checkable
Able to be judged without understanding the whole of it. The property that decides how much needs judgment at all.
constraint
The thing that limits output. Adding more of anything else does not help until it moves.
rework
Work that had to be revisited after somebody called it done.
vigilance
Staying alert for a rare problem across a long stretch of time.

Taken as already known, and so not defined here: agent, model, review, throughput, headcount. That list is a claim about who is reading, and it is printed so it can be argued with.