Aftersales KPIs for service departments, with formulas
The aftersales KPIs a service manager should track each week, with the formula for each, what moves it and how to read them in pairs.
- By
- Thilo Krause
- Published
- Reading time
- 8 min read
The aftersales KPIs worth tracking in a service department fit on one page. For the workshop's output, track hours sold per order, technician productivity and efficiency, and loading. For the health check, track the VHC rate and the red and amber work identified versus sold. For speed, track the response time at each hand-off and the invoice lead time. For quality and the customer, track first-time fix, the no-show rate and customer satisfaction.
Each has a formula you can calculate from your own orders, and each points at a different person or step when it moves. Below are the formulas, what moves each group, and a weekly routine for reading them together.
Aftersales KPIs for service departments on one page
| KPI | Formula | Moved mainly by |
|---|---|---|
| Hours sold per order | Sold labour hours ÷ orders | Write-up, health check, approvals |
| Productivity | Clocked hours on jobs ÷ attended hours | Dispatch, parts, waiting |
| Efficiency | Sold hours ÷ clocked hours, on clocked jobs | Skill, tools, labour times |
| Loading | (Booked + carried over) ÷ sellable hours | Booking, carry-over |
| VHC rate | Orders with a submitted check ÷ orders | Workshop routine |
| Red and amber sold | Items approved ÷ items found | Evidence, pricing speed |
| Response times | Minutes from one event to the next action | Hand-offs |
| First-time fix | Orders without a comeback ÷ orders | Diagnosis, final check |
| No-show rate | No-shows ÷ appointments due | Reminders, booking |
| Invoice lead time | Final check passed to invoice issued | Drafting, review |
| Customer satisfaction | Survey score, for example NPS | Promise times, hand-over |
Calculate all of them over the same 30 days, long enough to smooth out one bad Monday and short enough to show last month's change. Where the formula allows it, split the number per advisor and per technician. A workshop average can hide the one person who needs help.
Before you compare any of these with a figure from a manufacturer, a consultant or your DMS, check the definition. The same name covers different ratios. Prompt CRM calls clocked hours on jobs ÷ attended hours productivity. The Alltrucks workshop network uses "productivity level" for time sold ÷ time available and says it should be at least 85 % (Alltrucks). Both are sound, and they give different numbers for the same workshop. A target from one source measured with the other formula tells you nothing.
Workshop output: hours sold, productivity and efficiency
Hours sold per order. Divide the labour hours you invoiced by the number of orders. It is the simplest measure of how much work each visit brings in. A thorough write-up raises it, and so do health check findings that get sold. Advisors who trim labour time on the invoice to keep a customer happy lower it. Split it by payer: customer, warranty and internal work move for different reasons.
Productivity. Attended hours are the hours a technician is on site, after breaks and absences. Productivity is the share of those hours clocked on jobs. The gap is everything else, such as waiting for allocation, a part or an approval, fetching cars and cleaning the bay. Most of it sits outside the technician's control. A low productivity figure usually points at dispatch, at the parts desk or at an advisor who has not priced the findings yet.
Efficiency. Divide the hours sold on the jobs a technician clocked by the hours they clocked on those jobs. Above 100 % means the job took less time than you sold. Skill, tools, access to repair information and the labour times on your menus move it. Count only clocked jobs. If a job sold at two hours has no clocking, leaving it in the sold hours pushes efficiency up and hides the missing clocking.
Take a workshop with five technicians over four weeks. All numbers are invented and rounded. The technicians attend 800 hours and clock 680 hours on jobs, so productivity is 85 %. The jobs they clocked were sold at 714 hours, so efficiency is 105 %. The workshop closed 420 orders, so hours sold per order is 1.7. Multiply productivity by efficiency and you get 89 %. For every hour a technician spends on site, the workshop sells 0.89 hours.
Loading. Add the hours booked for a day to the hours carried over from the day before, and divide by that day's sellable hours. Look at it for each of the next ten working days, every morning. It is the only KPI on this page that looks forward, and it tells you whether tomorrow's promise times are realistic before the cars arrive.
Health check KPIs: checks done and work sold
The VHC rate is the share of orders with a submitted health check. Decide once which orders are eligible, for example everything except a tyre swap, and keep that rule. A low VHC rate is a workshop routine problem. Cars leave without a check because the technician ran out of time or nobody asked for one.
Red and amber identified versus sold shows what happens after the check. Count the red (safety) and amber (due soon) items the technicians found and the ones the customers approved. Count them twice, by number and by value. A workshop that sells many wiper blades and few brake jobs looks good by count and poor by value. A low red sold rate points at the evidence and at the advisor's explanation: the photo, the video and the one sentence the customer reads. It also points at speed. Findings that wait an hour for a price reach a customer who is already in a meeting.
Amber sells less on the day, and that is expected. Many customers postpone work that is due soon. The postponed items need a follow-up date on the vehicle record, or they are gone. The digital vehicle health check article covers the rating rules and a worked example of found versus sold.
Speed: response times and invoice lead time
Response times measure the hand-offs between roles. Take the four that cost the most when they stall:
- Check-in to the first job line allocated to a technician.
- Health check submitted to estimate sent to the customer.
- Estimate sent to the customer's decision.
- Final check passed to invoice issued.
For each, set a target time and report two numbers: the median and the share of cases within target. Use the median, because one car that waited overnight drags the average up for the whole month. When a response time slips, look at who owns that step and what else landed on their desk at the same hour. The service advisor workflow article goes through these hand-offs step by step.
The last one on the list, the invoice lead time, deserves its own line on the report. When it runs long, customers wait at the desk and the back office chases missing lines the next day. Typical causes are job lines never closed, an open warranty decision and a payer split nobody set up at check-in. If you invoice business customers, also count e-invoices that come back rejected. For German workshops, the XRechnung article explains which formats apply and when.
Quality and the customer: first-time fix, no-shows and satisfaction
First-time fix. Divide the orders without a comeback for the same concern within a fixed window by all orders. Pick the window once, for example 30 days. The hard part is counting comebacks. Ask at every check-in whether the visit is about a concern from a previous visit, and if it is, record a reason code and link the two orders. Diagnosis time, the write-up and a final check that tests the repair move this number. A workshop that pushes efficiency without watching first-time fix ends up doing the same repair twice and giving the second one away.
No-show rate. Divide the appointments not arrived by your cut-off time by all appointments due that day. Track the hours lost as well, because one missed timing belt costs more than three missed oil services. Confirmations, a reminder a day or two before and an easy way to rebook move it.
Customer satisfaction. Send a short survey soon after the hand-over. If you use the Net Promoter Score, Bain & Company defines it from one question, how likely the customer is to recommend you, on a scale of 0 to 10. Scores of 9 and 10 are promoters, 0 to 6 are detractors, and the NPS is the percentage of promoters minus the percentage of detractors (Bain & Company). The score matters less than what happens after a low one. Someone should call that customer within a day. In Prompt CRM, the survey goes out 24 hours after hand-over and a low score creates a follow-up task for the service manager.
A weekly routine for reading the numbers
Read the KPIs in pairs, because one number alone rarely tells you who should act:
- Productivity with efficiency. Productivity rising while efficiency falls often means waiting time has moved onto job clockings.
- VHC rate with red sold rate. A VHC rate that jumps while the sold rate drops suggests checks done to meet a target, with weak photos and few findings that convince anyone.
- Hours sold per order with response times. If pricing slows down, hours sold per order usually follows a week or two later.
- Efficiency with first-time fix. High efficiency with rising comebacks points at rushed repairs.
Then keep a fixed half hour every Monday:
- Compare each KPI with its own last four weeks, and ignore single-week jumps.
- Pick the one that moved most in the wrong direction.
- Open five orders behind it and read what happened on each.
- Give one person one change to make this week, and check the same number next Monday.
Prompt CRM's service manager report shows productivity, efficiency, hours sold per order, the VHC rate, the video rate, red and amber findings identified versus sold, and response times against the service levels, over 30 days. First-time fix is not in the report yet, so count comebacks with a reason code as described above.
Start with the KPIs you can calculate from your own orders this week, write the formula next to each one and keep the definitions fixed for at least a quarter. A number whose definition changes every month cannot tell you whether last month's change worked.