Comeback work: the job nobody pays for.

Count it before you argue about it

Comeback work is almost never in the numbers, because it is rarely invoiced and therefore rarely recorded anywhere the reporting can see.

That means most workshops are debating a problem whose size nobody knows. The fix is unglamorous: log every return visit related to previous work, with the hours spent and the parts used, whether or not anyone was charged.

A quarter of that data usually settles the argument. Comebacks are almost always concentrated in a small number of job types rather than spread evenly, which makes them fixable rather than a general standard-of-work problem.

The common causes are boringly specific

When the log exists, the same handful of causes account for most of it.

• A part that was not the cause, fitted because the diagnosis stopped at the first plausible answer.
• A related item not replaced at the same time, which then failed and looked like the same fault.
• Something disturbed during the job and not reassembled correctly, which is the category customers find most damaging to trust.
• A poor-quality part chosen on price for a job where it was the wrong economy.
• The customer was never told what to expect, so normal behaviour after a repair reads as a failure.

That last one is worth separating from the rest, because it is not a technical comeback at all and it is fixed with a sentence rather than a process.

Diagnose to a cause, not to a plausible part

The most expensive comebacks come from replacing something reasonable that was not actually at fault.

It happens under time pressure, and it is rational in the moment: the part is a common cause, it is in stock, and the alternative is more diagnostic time the customer may not want to pay for. The cost lands later, as a second visit, a second part and a customer who is now paying twice for one fault.

The defence is charging properly for diagnosis rather than absorbing it. A workshop that gives away diagnostic time is under pressure to guess, and guessing is what produces comebacks.

Set expectations at handover

A meaningful share of returns are vehicles with nothing wrong with them.

New brakes need bedding in, a replaced component can smell as it heats for the first time, and a fault that was masking another one means the second becomes noticeable afterwards. None of that is a failure, and all of it reads as one to a customer who was told nothing.

Thirty seconds at handover removes most of these. It also makes the genuine comebacks easier to distinguish, because the customer describing a problem has already ruled out the normal things you warned them about.

Handle the real ones generously and quickly

When a comeback is genuine, arguing about it is more expensive than fixing it.

The cost of the rework is usually smaller than the cost of a customer telling people about how it was handled, and a workshop that fixes its mistakes visibly and without friction often ends up with more trust than one that never made the mistake.

What that requires is a rule agreed in advance, so front-of-house does not have to negotiate: comebacks get priority in the schedule, they are diagnosed by someone other than the person who did the original job where possible, and the customer is not asked to justify themselves.

Feed it back into how the work is done

A comeback log that nobody reads is just a longer record of the same problem.

Reviewing it quarterly, looking for the concentration rather than the individual cases, usually produces two or three specific changes: a job type that needs more time in the schedule, a part that should not be bought on price, or a step that needs to be on a checklist rather than in someone's memory.

That is the whole return on measuring it. The point was never the number, it was knowing which two or three things to change.