Rework

The largest hole in most small operations, and it never appears on a single report.

A callback produces no invoice, so the hours vanish into the general cost of doing business, and the man carrying them experiences it as being busy.

What it actually costs

Put numbers on it, because the abstraction produces nothing.

Nine thousand dollars, and that is only labor. It does not count materials, fuel, the schedule slot that could have held a paying job, or the customer who now tells people the work had to be redone.

Add the opportunity cost and the real figure is closer to twenty thousand.

Now compare that to what a man would do for a twenty-thousand-dollar increase in revenue. He would work weekends for it. He will not spend four hours writing a checklist that produces the same result, because one is visible and the other is invisible.

Where it actually comes from

Four sources, and only one is skill.

Unclear scope. The customer expected something that was never in the quote. Not a workmanship failure. A proposal failure, fixed on paper before anybody picks up a tool.

And actual skill gaps, which are the least common cause and the first one every owner assumes.

Three of the four are systems problems.

Which means most rework is not a people problem and cannot be fixed by hiring better, and men spend years trying anyway.

The thirty-day log

The entire intervention, and it is one line per event.

Date. Job. What happened. Why.

Thirty days. Four to fifteen entries for most operations.

At the end they will cluster into two or three causes, and both of those causes will be fixable with a checklist and a paragraph in the quote template.

Cutting rework in half is a margin improvement of two to four points for most service operations, and unlike a price increase it requires no conversation with anybody.

It is the least glamorous profit available and it is sitting on the floor of your business right now, uncounted.

A real log, thirty days

Here is what the exercise actually produces. One line per callback, no analysis while writing.

Nine callbacks. Twenty-two hours. Roughly $1,150 in labor alone.

Now sort by cause.

Four causes. All four fixable in under three hours of writing.

And notice the pattern nobody would have seen from memory. Three separate customers expected something that was never in the quote, and none of them were unreasonable. That is not three difficult customers. That is one missing line on a template, and the owner would have described those three as bad luck.

Notice also Whitfield and Kaufman and Reyes appear twice each. A callback that produces a second callback is a signal that the first visit was also rushed, which is the same cause showing up as two events.

Nine entries. Thirty days. Three hours of fixes against roughly $13,800 a year.

The culture problem underneath

If reporting a callback gets somebody in trouble, callbacks stop being reported.

They do not stop happening. They get handled quietly, between the crew, and the pattern that would have shown you the cause never surfaces.

So the log has to be explicitly blameless, and it has to be blameless the first time, in front of everybody, when the temptation to react is highest.

I'm not looking for who. I'm looking for why. Write it down and keep going.

A man who reacts badly to the first entry has ended the exercise and will not know it for six months.

The three fixes that cover most of it

Itemize exclusions on every quote. One line. Does not include: tile repair, painting, relocating the existing fixture.

Twenty seconds to write and it prevents an entire category of dispute, which is scope creep arriving as a callback.

And photograph the finished work. Every job, before departure. Three photographs.

That does two things. It creates a record when a customer reports something later, and more importantly, a man who knows he is photographing the finish looks at the finish differently, which is most of the benefit.

Drawn from Built to Last

disciplinefocussystemstime