Auto shop bottleneck detection: find where cars sit and why

A tech finishes an inspection at 9:40. The estimate does not start until 10:10 because the advisor is on the phone. The customer approves at 1:00 because the text went out after lunch. The part arrives tomorrow. The car has now been in your building for two days, and the tech only worked on it for about an hour.

Nobody in that story was slow. The car was waiting. Auto shop bottleneck detection is the habit of finding where cars wait, how long, and for what, so you fix the real constraint instead of pushing everyone to move faster.

What a bottleneck is, in a shop

The Lean Enterprise Institute describes the theory of constraints as an approach, developed by Eliyahu Goldratt and popularized in his 1984 book The Goal, that focuses on removing or managing the constraints that prevent an organization from reaching its goals. The practical reading for a shop: the whole building can only complete as many cars as its slowest stage allows. Speeding up any other stage changes nothing.

In a shop, the slowest stage is rarely the wrenches. It is more often one of these:

• Estimate creation. Inspection is done, but nobody has started the quote.
• Customer approval. The quote is out and the car sits on a lift or in the lot.
• Parts. The job is approved and the part is not in the building.
• Bay or tech availability. Work is approved and ready, but the right lift or the right certified tech is busy.
• Final inspection and pickup. The car is done and the customer has not been told.

Why delays hide

AutoVitals, a shop-software vendor, put it plainly in a March 2026 post: delays rarely come from one big problem, they show up as small waits such as a car sitting before its estimate or a technician pausing for an approval. A shop owner walking the floor sees busy people and assumes the process is fine.

One more trap is the owner. Shop4D, another shop-software company, lists the owner as a bottleneck when every decision needs their sign-off. If your advisors cannot discount a part, add a sublet or approve a goodwill fix without calling you, your approvals stage is you.

A two-week method

You do not need software to start. You need a clock and a consistent status.

1. Define the stages. For example: checked in, inspecting, estimate in progress, waiting on approval, waiting on parts, in repair, final check, ready for pickup. AutoVitals recommends a clear status on every repair order, because unclear status is where jobs go missing.
2. Record time in each stage. Most shop systems stamp status changes. If yours does not, use a whiteboard and initials for two weeks.
3. Tag every wait with a cause. Not "slow," but "no estimate started," "customer has not answered," "part ordered, no ETA," "no free lift."
4. Add up by cause. At the end of two weeks, total the hours by cause. One or two will dominate.
5. Fix the biggest one only. Then measure again. Fixing three things at once tells you nothing about which worked.

Walk a few orders from check-in to pickup by hand before you trust the totals. If the numbers disagree with what your advisors tell you, ask why.

Common causes and what each fix looks like

Waiting on estimates. Standardize inspection findings so an estimate can be built quickly, and set a target time from inspection complete to estimate sent. AutoVitals also suggests standardizing inspections and speeding approvals as the main levers.

Waiting on approval. Phone tag is the usual cause. A text with photos and a link sent while the car is still on the lift, plus a rule for a second contact attempt, closes most gaps. See repair status automation for the messages that go with it.

Waiting on parts. Give one person ownership of parts status, and record an ETA on every ordered part. A job with no ETA is the one to chase today. Order commonly needed parts early for jobs you expect to be approved, and accept that you will return some.

No lift or tech available. Check whether the constraint is lifts, or one certified tech who gets every electrical job. Technician time is a real constraint: the Ratchet+Wrench 2024 industry survey found 27% of shops report technician efficiency below 80%, and 18% do not track it at all. Low efficiency can come from waiting, not from slow work, so look at the wait causes before you coach the tech.

Where AI helps

AI does not remove a bottleneck. It makes one easier to see, sooner.

• Watching age. Flagging every repair order that has sat in a stage longer than your own threshold, so the advisor does not need to scan the board.
• Spotting patterns. Shopmonkey's 2024 post on AI lists identifying repairs that take longer than expected as a use of analytics. A weekly summary that says "approvals took longest on Tuesdays" is useful.
• Drafting the next step. A follow-up for a stale approval, written for a person to review.
• Saying what it does not know. A delayed part with no ETA should show as unknown, not as a guess.

What stays with a person

Deciding which wait is worth fixing, changing how work is dispatched, and talking to a customer about a delay. AI reads timestamps. If your team changes statuses late, the data will point at the wrong stage.

Where RevvOS fits

RevvOS's daily operations view includes a job board showing the stage, age, assigned person and problems for each repair order, and a short list of stale estimates, aging repair orders, delayed parts and callbacks owed. When a part has no delivery time, it says so. It works alongside your shop system and prepares next steps for a named person to review. It does not measure bottleneck savings for you, and we publish no figure for it.

Related reading: auto repair operations software AI and technician productivity AI auto repair. To pick one workflow and look at it with your own repair orders, start with the Multi-Shop Operating Gap Assessment.