Mechanic shop software: the approval queue is the real bottleneck

A collision shop waits on insurers. A tire shop waits on a rack. A mechanical shop waits on customers — and it is the least visible queue in the trade, because a car waiting on a customer looks exactly like a car waiting on anything else.

The shape of the problem

The sequence is always the same. A car comes in. It gets inspected and diagnosed. An estimate goes out. And then nothing happens, for a length of time entirely controlled by somebody who is at work and not thinking about their car.

During that time the car is occupying a space, the job is open, and the shop's capacity is committed to work that may not be approved. If three cars are in that state at once, a meaningful fraction of the shop is doing nothing and it does not look like it.

Why it hides

On most boards, "waiting on approval" is not a stage. The car sits in whatever column it was in — diagnosed, or estimated, or just in progress — and its real state lives in the advisor's memory or a text thread on their phone.

That has two consequences. Nobody except that advisor knows the car is stuck, so nobody else can help unstick it. And when the advisor is off, the car is invisible entirely.

The fix is to make the wait a place rather than a fact. Estimate and waiting-on-approval as one column group, two lanes: the estimate being written, and the estimate sent and unanswered. A car in the second lane is visible to everybody, and how many cars are in it is the number that predicts the week.

Approvals and parts are different emergencies

Two cars sitting untouched can look identical and need completely different work.

One is waiting on a supplier. The action is to chase the supplier, or find another one, or decide the job waits.

One is waiting on a customer. The action is a phone call, and it is a phone call that gets easier the sooner it happens — a follow-up on day one is a helpful nudge, on day four it is an awkward conversation about a car nobody has touched.

If both cars sit in the same column, whoever looks at the board has to ask which is which before they can do anything. Separate columns mean the board tells you which call to make before you walk over.

Making the wait shorter

Once the queue is visible, most shops find the same three things shorten it.

Send it faster. The gap between diagnosis and the estimate going out is entirely yours. It is often longer than anyone realises, because it is nobody's specific job.

Make it answerable. An estimate that requires a phone call during business hours gets answered slower than one a customer can approve from a phone at nine in the evening. Most people are not available at the times a shop is.

Follow up on a schedule, not on a feeling. A card that has been in waiting-on-approval for more than a day is a follow-up. Not because anything is wrong, but because the customer has forgotten, and a reminder is a favour.

None of these need new software. They need the queue to be visible enough that somebody owns it.

What to check in a tool

If you are evaluating software for a mechanical shop, the approval queue is the thing to test, because it is the thing most tools handle worst.

Can waiting-on-approval be its own column, distinct from waiting-on-parts and from active work? Can anybody see how long a car has been in it, without opening the car? Can the estimate go out in a form the customer can answer without calling back during business hours? And when they answer, does the answer land somewhere attached to the job rather than in one person's messages?

A tool that gets those four right will do more for throughput than any amount of scheduling cleverness. The work was never the constraint.