Garages that are busy and not making money are rarely losing it in one obvious place. If they were, somebody would have found it.
It goes in small amounts, on ordinary jobs, in ways that never present as a problem, because each individual instance is too small to notice and there is no report that shows the total.
Parts fitted and never charged
The most common one, and the easiest to fix.
A clip breaks on removal. A technician takes a replacement from the stores, fits it, and gets on with the job. It is a four-pound part and stopping to write it down feels disproportionate.
Multiply by a few a day, across the technicians, across the year. It is not four pounds. And the cost is not only the part. It is that your stock figures are wrong, so reordering is wrong too.
This is a workflow problem, not a discipline problem. If recording the part means walking to the office, it will not happen. If it means three taps on the phone already in their hand, it will.
Authorised extras that never reach the invoice
Worse, because the amounts are larger.
The customer approves two extra items on the phone. The work gets done. At invoicing, one of them is missing, because the approval happened in a conversation and the invoice is built from the job card.
Nobody notices. The customer certainly does not ring to say you undercharged them.
The fix is structural: the approval and the invoice have to come from the same record. When a customer approves a health-check item and that approval carries through to the invoice, the item cannot be lost between the two, because there is no gap for it to fall into.
Time nobody recorded
A job quoted at two hours that took three.
If the extra hour is not recorded, two things follow. The invoice is short by an hour. And: worse in the long run. Your data now says that job takes two hours, so you will quote it at two hours again, and lose the same hour every time.
That is the compounding one. Underbilling once costs you an hour. Recording it wrong costs you an hour on every one of those jobs, indefinitely, until somebody notices the job type is unprofitable and cannot work out why.
This is why quoting from standardised repair times rather than memory matters. A published time is a benchmark you can compare actuals against, which is what makes the gap visible at all.
The comeback nobody counted
A car returns for something related to the previous job. It gets sorted, no charge, quite right.
But the hour it took is not on any job, so the workshop's productive hours look better than they were, and the original job looks more profitable than it was. The cost is real; it is just recorded nowhere.
Comebacks are worth recording precisely because they are not charged. Otherwise they are invisible, and a job type with a comeback rate is indistinguishable from one without.
How to find yours
You do not need a project. You need three numbers for one month.
- Parts issued from stock versus parts invoiced. The gap is the parts leak.
- Hours recorded against jobs versus hours paid. The gap is unrecorded time, comebacks and genuine downtime: worth separating.
- Approved health-check items versus invoiced items. The gap is authorised work that was never billed.
Any accounting for these is better than none. The first month is usually uncomfortable and always instructive.
The reason these leaks persist is not that garages are careless. It is that each one is small enough to be beneath notice at the moment it happens, and there is no place where they add up. Making them add up somewhere is most of the fix.
Do not fix this with a policy
The instinctive response is a rule: everybody records every part, no exceptions.
It will hold for about three weeks. Rules that ask people to do something inconvenient at the worst possible moment: hands dirty, car on the ramp, customer waiting: lose to the moment every time. That is not indiscipline, it is arithmetic.
What holds is making the right action the easy one. Recording a part from the bay in three taps beats a policy about walking to the office. An approval the customer taps beats a rule about writing down what they said on the phone.
If a leak keeps reappearing after you have fixed it, the fix asked someone to be more careful rather than making carefulness unnecessary.
The one number to watch
If you track nothing else, track sold hours against paid hours, monthly.
It is a single figure, it is hard to argue with, and almost every leak in this article shows up in it eventually. Time not recorded pushes it down. Comebacks push it down. Jobs quoted short push it down.
You do not need a target. You need a trend. A number drifting downwards month on month is telling you something is wrong well before the accounts do.
