Bay throughput: the metric that separates good service departments from great ones

What bay throughput actually means

Bay throughput is cars completed per bay per shift. Not RO count. Not labor hours. Not gross. Just how many cars actually rolled off a given bay during a given shift.

The reason it's the right metric is structural. Your fixed costs in fixed ops are tied to bays and shifts: the building, the lifts, the techs you have on payroll, the supervisor's salary. Revenue is variable. Throughput is the bridge between the two. Two stores with the same number of bays and the same labor rate can have wildly different P&Ls because one of them moves seven cars per bay per shift and the other moves five.

Most fixed ops directors track ROs, hours per RO, and gross profit. Those are lagging metrics. Throughput is leading. If your throughput goes up, your ROs and your gross will follow next month. If your throughput is flat, your ROs are flat — even if your CSI scores look great.

Industry benchmarks: express vs full-service vs quick lane

"Good throughput" depends entirely on what kind of bay you're running. A general service bay handling diagnostics and a quick-lane bay running oil-and-filter are not the same animal and shouldn't be benchmarked against each other. Here's what the data looks like across categories. These ranges come from talking to ~50 fixed ops shops and from public 20-group composites; treat them as ballpark, not gospel.

Bay type	Typical job	Avg job time	Cars per bay per 8hr shift	Best-in-class
Quick lane / Express oil	LOF, tire rotation, multi-point inspection	20-30 min	10-14	16-18
Express service	LOF + minor menu items, brakes, batteries	40-60 min	6-8	9-10
Main shop / Full service	Diag, drivability, larger repairs	1.5-3 hrs	3-4	5-6
Heavy line / Internal	Used car recon, transmission, engine	4-8 hrs	1-2	2-3

Two patterns to notice. First, the gap between average and best-in-class is roughly 30-40% in every category. Best-in-class isn't running a different shop; they're running the same shop with fewer dropped balls. Second, the highest-leverage category is express. A quick-lane bay moving 18 cars instead of 12 generates 50% more revenue from the same fixed cost. A heavy-line bay moving 2 cars instead of 1 doubles its output but the absolute revenue gain is smaller in dollars.

If you're prioritizing where to invest time and process change, start with express and quick lane. The leverage is highest, the changes are easiest, and the visibility is best.

The four levers that move the number

You'll see a hundred articles online listing "20 ways to improve service efficiency." Most of them are noise. There are really only four levers, and every successful improvement is some combination of them.

Lever 1: Routing (which car goes to which bay/tech)

This is the lever almost nobody manages explicitly. Most stores route on advisor habit and tech tenure: senior tech gets the gravy work, junior tech gets what's left, all of it filtered through whichever advisor wrote it. The result is uneven utilization — a couple of techs running hot, a couple of techs idling, and the average looks "fine."

Smart routing assigns by fastest qualified tech currently available for this service type. The qualified part matters: you don't send a complex transmission to a B-tech because they're free. But within the pool of qualified techs, the fastest available wins. We have data on this in the routing section below.

Lever 2: Downtime (the gaps between cars)

Tech downtime between jobs is the single largest source of throughput loss in most shops. The bay is open, the tech is qualified, the next car is sitting in the lot — but nobody has told either of them that. Average between-job downtime in stores we've measured runs 12-18 minutes per car. That's two extra cars per bay per shift, gone.

Downtime collapses to under 4 minutes when (a) the queue is visible to the tech without having to ask the advisor, (b) the next car is auto-assigned the moment the previous job closes, and (c) the customer is already in the building or has confirmed arrival.

Lever 3: Parts staging

Tech grabs the keys, lifts the car, opens the hood, walks to the parts counter for an air filter, waits 4 minutes for someone to find it, walks back. Repeat 8 times a day per tech across 9 techs and you've burned an hour per tech per day. Pre-staged parts — for menu items especially — recover that hour with zero capital expense.

The fix is workflow, not software: when a customer is added to the queue, a parts pull ticket prints in parts. When the tech is assigned, the parts are already on the bay cart. You can do this with a clipboard. You do it faster with a queue tool that fires a webhook to parts the moment a service is selected.

Lever 4: Advisor handoff

The handoff between "tech finishes work" and "advisor calls customer" is where minutes disappear. Tech wraps the RO. RO sits in the advisor's queue. Advisor is on a phone call. RO sits. Customer is on the bench, getting impatient. Advisor finally gets to it 12 minutes later.

This is fixable with two changes. First, the tech-to-advisor signal needs to be active, not passive: an alert, not a row in a list. Second, customer notifications shouldn't wait for the advisor's call — the SMS "your car is ready" can fire automatically the moment the tech marks the RO complete, and the advisor catches up at their pace.

How to measure throughput without expensive tools

You don't need an MES system. You need a clipboard, a stopwatch, and one Saturday.

1. Pick one bay. Ideally express or quick-lane.
2. Mark four timestamps for every car for two weeks:
   • T1: Car enters the bay (lift starts to rise)
   • T2: Tech actively starts working
   • T3: Tech finishes work, RO closed
   • T4: Car leaves the bay
3. Calculate three intervals:
   • T2 − T1 = "ramp-up" (parts grab, RO read, lookup)
   • T3 − T2 = "wrench time" (actual work)
   • T4 − T3 = "wrap-up" (paperwork, signal advisor, car out)
4. Add up cars completed per shift. Compare to the benchmark above.

The pattern that almost always shows up: wrench time is fine. It is consistent, professional, and matches the SRT. The variance is in ramp-up and wrap-up. Stores that are below benchmark almost never have a wrench-time problem; they have a between-jobs problem. That is great news, because between-jobs is exactly the problem your queue and routing system can solve without changing how techs work.

If you want to skip the clipboard, any modern queue platform records these timestamps automatically when techs start and complete their assignments. The point is not which tool — the point is having the timestamps.

Smart vs round-robin auto-assignment

The biggest single throughput unlock we've seen — and we've watched a lot of changes — is moving from advisor-discretion or round-robin routing to smart routing. Round-robin is "next car goes to next tech in line, regardless." Smart routing is "next car goes to the fastest qualified tech currently available, with a tiebreaker on workload balance."

Here's what the data looked like at a real Subaru express service center we worked with (numbers anonymized but real). Three express bays, four techs, average mix of LOF, tire rotation, brake jobs, batteries.

Metric	Round-robin (baseline)	Smart routing	Change
Avg cars per bay per 8hr shift	6.4	7.8	+22%
Avg between-jobs idle time	14 min	5 min	−64%
Tech utilization variance (std dev)	22%	9%	tighter
Customer wait (P50)	34 min	26 min	−24%
Customer wait (P90)	78 min	49 min	−37%
CSI 'wait time' sub-score	4.31/5	4.62/5	+0.31

The big P90 movement is the most important number in that table. P50 is the median customer experience; P90 is the bad customer experience. Round-robin sends a brake job to a tech who happens to be next in rotation but who's slower at brakes than the tech standing right next to them. The slow tech, doing the wrong job, becomes a tail-risk that pushes P90 wait into "I'm never coming back" territory.

Smart routing solves this without speeding anyone up. The faster tech does what they're already fast at; the slower tech does what they're qualified for and not bottlenecking on. Total wrench-time is the same. The variance collapses, and so does customer pain.

A 90-day plan to lift throughput 15-20%

This is achievable in most stores without adding bays, hiring techs, or replacing your DMS. It requires consistent management attention and a willingness to measure.

Days 1-14: Measure

• Pick your most-used bay tier (usually express).
• Capture the four timestamps for every car for two weeks.
• Calculate cars/bay/shift, ramp-up, wrench, wrap-up.
• Identify the top three bottlenecks. They will almost certainly be: parts grab time, advisor handoff lag, and routing imbalance.

Days 15-30: Fix parts staging

• Define the top 10 menu items by volume.
• For each, list the SKUs needed.
• Pre-stage carts: when a customer is queued, parts pulls and stages on the cart before the tech is assigned.
• Measure: ramp-up time should drop 4-7 minutes per job.

Days 31-50: Fix routing

• Pull tech-by-service completion times from the last 90 days. (Most queue platforms do this; if not, manual capture works.)
• Build a simple matrix: tech rows, service columns, average minutes in each cell.
• Switch from round-robin to fastest-qualified routing. Communicate why to the team — this is essential, see our article on data-driven tech performance.
• Watch P90 wait drop by 20-30%. Watch CSI track up.

Days 51-70: Fix the handoff

• Active alert from tech to advisor when RO is wrapped (not a passive list).
• Auto-fire customer SMS the moment the RO is marked complete.
• Advisor follows up at their pace; the customer is no longer waiting on the advisor's phone schedule.
• Measure: wrap-up time should drop 6-10 minutes.

Days 71-90: Tighten and repeat

• Re-measure throughput. You should be 15-20% above day-1 baseline.
• Hold a 30-min review with techs. Show them the chart. They are the ones who made it happen; tell them.
• Pick the next bay tier (main shop or quick lane). Repeat the cycle.

Honest tradeoffs and what can go wrong

This is the part most articles skip. Here's where these changes can backfire and how to keep them from doing so.

Speed at the cost of quality

If you incentivize throughput in isolation, techs will start cutting corners — skipping torque checks, half-doing inspections, missing recalls in their hurry to close the RO. The fix is to pair throughput with come-back rate and CSI. We cover this in detail in why tech performance data should be running your bay assignments. The short version: never measure speed without measuring quality.

The senior tech revolt

Smart routing means your most senior tech doesn't always get first pick of the gravy. They will notice. They will complain. The way through is transparency: show them the data, show them their own numbers (which are usually excellent for their specialties), and tie payouts to actual flag hours rather than first-pick advantage. Most senior techs come around once they realize they're not losing money — they're losing the politics of who got which RO.

Burned-out fastest tech

Without workload balance in the routing logic, your fastest tech eats every easy job and burns out by Wednesday. Always combine speed-based routing with a workload cap or fairness term.

Throughput tunnel-vision

If throughput becomes the only metric the manager talks about, advisors stop selling and start sprinting customers through. CP per RO drops. The fix is balance: throughput and CP/RO should be reviewed side by side every Monday. We sketch what that looks like in 5 metrics every service manager should review every Monday morning.

Tools that promise more than they deliver

A queue platform can't run your shop for you. It can capture timestamps, surface routing recommendations, and remove paperwork friction. It cannot make a parts manager pre-stage parts. It cannot make an advisor stop chatting on the phone for 12 minutes. The throughput gain comes from the management layer using the data the tool provides. If you buy a tool and don't change the workflow, throughput won't move.

Why this metric, why now

Customer-pay volume is under pressure across the industry. EVs need fewer scheduled services. New-car warranty periods are getting longer. The traditional "bring 'em in for a 30K service" rhythm is getting weaker every year. The stores that thrive in the next five years will be the ones that wring more revenue out of the visits they already get — through recall capture, menu attach, and more cars per bay per shift. Throughput is no longer a nice-to-have. It's the line that separates stores that grow from stores that contract.

Related reading: Why tech performance data should be running your bay assignments · The psychology of waiting