HPRO - Hours per repair order
Hours per repair order explained: what HPRO measures, and what it does not
Hours per repair order, shortened to HPRO, is the average number of labour hours sold on each repair order in a period, calculated as total sold hours divided by the number of repair orders closed. It measures how much work each vehicle visit produces, not how fast the workshop gets through it.
Written as HPRO, hours per RO, or labour hours per RO, and spelled "labor" in North American reporting. Every fixed ops dashboard carries it, and almost every one of them carries it blended, which is where the trouble starts.
What does hours per repair order actually measure?
The sold hours on a repair order are the hours a customer agreed to pay for. So HPRO is a presentation and approval metric sitting in a throughput dashboard, and that is why it behaves oddly next to the numbers around it.
Two workshops with identical technicians and identical labour times will report different HPRO. One inspects every vehicle, presents what it finds, and gets a decision. The other writes up what the customer asked for at the desk and hands the keys back. The technicians are not the variable. The conversation is.
This is what separates HPRO from technician efficiency. Efficiency asks how quickly sold hours get completed. HPRO asks how many were sold in the first place. A shop can run 130% efficiency on 1.2 hours per repair order and still have empty bays by Thursday.
How is HPRO calculated?
HPRO = total labour hours sold in the period ÷ repair orders closed in the period
Three parameters have to be fixed before the number transfers between two people:
- Which repair orders count. All of them, or customer pay only? Including internal and express work changes the answer materially, for reasons covered below.
- Sold hours, not clocked hours. The numerator is billed labour time from the time schedule. Clocked hours produce a different metric entirely and a lower one in a fast shop.
- Closed or opened. Count repair orders on the same basis every month. Long jobs straddling a period boundary will otherwise move the figure on their own.
Hours per repair order: a worked example
The average US franchised new-vehicle dealership wrote 16,252 repair orders in 2025 (NADA Data 2025, Annual Financial Profile of America's Franchised New-Car Dealerships). Call that roughly 1,354 a month. If that shop sells 2,300 labour hours in a month, HPRO is 2,300 ÷ 1,354 = 1.70.
Now add a quarter of an hour per repair order without booking a single extra car. At 1.95, the same 1,354 visits sell 2,640 hours, a gain of 338 hours. At an effective labour rate of 120 that is 40,560 in additional labour a month, before parts, from vehicles already in the building.
That arithmetic is the whole argument for managing HPRO ahead of repair order count. More cars needs more bays, more technicians and more marketing spend. More hours per car needs a better inspection and a better conversation.
Why the blended figure moves on its own
Repair orders are not one thing. Across all US franchised dealerships in 2025, NADA counted 112.78 million customer mechanical repair orders, 57.68 million warranty claim repair orders, 48.58 million internal repair orders, 34.10 million express service repair orders and 15.17 million service contract repair orders (NADA Data 2025).
Express and internal work carries small hour counts by design. Open an express lane and blended HPRO drops immediately, while total sold hours and total gross both rise. A manager reading only the blended line concludes the shop got worse in the month it got busier.
So report HPRO by category. Customer mechanical HPRO is the number worth managing. Warranty HPRO tracks the manufacturer's time allowances more than your own performance, and internal HPRO mostly reflects the used vehicle department's reconditioning policy.
How HPRO gets misread
- Read as a speed metric. It contains no information about how long anything took. That is efficiency, and the two numbers routinely move in opposite directions.
- Split visits. Opening a second repair order for the additional work approved during the same visit doubles the count and halves the average. The workshop did exactly the same amount of work.
- Clocked hours in the numerator. A common data entry error that quietly makes an efficient shop look weak.
- Chased through overselling. HPRO rises fastest when advisors sell work the vehicle does not need. It comes back as declined work next visit, a CSI score, and a customer at the independent down the road.
- Benchmarked against circulated ranges. Figures like 2.5 to 3.25 hours appear across fixed ops trade blogs, but without a published dataset or a stated definition behind them. NADA Data 2025 publishes repair order counts and sales per repair order, and no HPRO figure at all. Trend your own number by category instead.
- Averaged across advisors. The spread between the top and bottom advisor on the same drive is usually wider than the spread between two dealerships.
What actually moves it
Four things, roughly in order of how much they matter. Inspection completion, because work never found is never sold. How the finding is put to the customer, which is the whole subject of declined work and recommended work. Parts availability, because an approved job with no part on the shelf turns into a deferral. And workshop loading, because a bay booked solid with 0.4 hour jobs has no room for the 2.6 hour one.
The last practical note: HPRO is per repair order, not per vehicle and not per customer. A fleet customer bringing eight vans on eight repair orders counts as eight. Decide whether that matches the question you are asking before you put the number in front of anybody.
