Rework, Defects and the Cost of Doing It Twice
Work done again is effort spent with no output. It is usually the largest avoidable loss and the least measured.
Rework consumes capacity and produces nothing new. Most organisations do not measure it, and most that start are surprised by the figure.
The process issue in “Rework, Defects and the Cost of Doing It Twice” is easier to diagnose when effort can be connected to projects without confusing activity with value. Used for how to handle multiple clients, go to the website can add time and workflow context, while the process outcome remains the source of truth for whether the system improved.
What counts as rework
Something returned for correction.
For an independent perspective related to “Rework, Defects and the Cost of Doing It Twice”, consult the Microsoft WorkLab; it offers a useful external check on definitions, governance and the assumptions built into a proposed measure.
A ticket reopened.
A document revised because the brief was wrong.
A part remade.
A decision reversed because the information was incomplete.
And work abandoned part-way, which is the invisible category.
Why it is the best single measure
It is unambiguous: either it came back or it did not.
It is independent of the unit problem, because any item can be reworked.
It directly measures waste rather than inferring it.
And it is the counterweight that makes volume measurement honest, which the previous note argues.
Where it comes from
Unclear requirements, which is the commonest cause by a wide margin.
Missing information at the start.
Handoffs, where context is lost.
Rushing, which is frequently caused by the volume measure itself.
Each is a system cause rather than an individual one, which is the systems section's argument.
Measuring it
Count items that came back, as a proportion of items delivered.
By stage, where you can: rework found internally is cheaper than rework found by the customer.
And record the cause, in one word, chosen from a short list.
That cause field is where the improvement comes from, and it takes five seconds per item.
The ratio worth watching
Internal rework against external.
Finding problems yourself is good; the customer finding them is expensive.
A rising external proportion means checks are weakening, which is usually a consequence of pressure on volume.
What not to do with it
Attribute it to individuals.
Rework usually originates upstream of whoever did the work — in the brief, the handoff or the deadline.
Measuring it per person produces concealment, and concealed rework is both unmeasured and still expensive.
The honest conversation it enables
"A fifth of our capacity goes on doing things twice, and most of it traces to incomplete briefs."
That sentence is more useful than any productivity percentage, and it names an action.
Which is why rework is worth measuring even when nothing else is.
What to check
Do you measure rework at all?
Can you say what proportion of capacity it consumes?
Is the cause recorded?
And is internal rework distinguished from external?