Every dashboard in this industry can show you 50 numbers. Most of them go green while the business struggles. First-time fix looks great because techs close jobs as "complete" and open a new one for the return visit. Utilization looks great because drive time got logged as work. The owner stops trusting the report, and everyone goes back to gut feel. The fix isn't more field service metrics. It's fewer of them, calculated so they can't be gamed, and read together so one can't improve at another's expense.
Here are nine that actually move margin for a trades contractor. Each comes with its formula, how to measure it honestly, the trap to avoid, and what really moves it.
Nine metrics that move margin
The first four are about getting the right tech to the right job without wasting the trip. The next three are about what a tech's day produces. The last two are about the promises you make and the cash you collect.
1. First-time fix rate
Formula: jobs resolved on the first visit ÷ all jobs that needed an on-site visit.
Measure it honestly: a "return visit" is any visit to the same customer for the same problem within 30 days, whether or not someone opened a new job for it. If your system counts only reopened jobs, techs will learn to open new ones.
The trap: excluding "parts ordered" jobs as if they don't count. A missing part is exactly the failure this metric exists to catch.
What moves it: the right tech on the job, the right parts on the truck, and the equipment's history in the tech's hand before arrival.
For context: TSIA, which benchmarks mostly technology and equipment service organizations, reported an industry median of 87% of on-site incidents resolved in one visit (2021). Trades work differs. Use it as a reference point, and set your target from your own baseline.
2. Callback rate
Formula: jobs that generate a return visit for the same issue within 30 days (or your warranty window) ÷ completed jobs.
Measure it honestly: track it per tech and per job type. Otherwise you'll spend the fleet average on everyone when two techs and one job type are the actual problem.
The trap: confusing callbacks with first-time fix. First-time fix is about finishing on the first visit. A callback means the job was "finished" but the problem came back.
What moves it: job checklists, photos of the finished work, and coaching the specific techs whose numbers stand out.
3. Response time against the SLA target, by tier
Formula: calls answered inside the SLA response target for that customer's tier ÷ all calls with a response target, reported separately for each tier.
Measure it honestly: start the clock when the customer asked, not when the dispatcher entered the job. Define "response" once, the way your contracts do (a tech on site, or a confirmed dispatch), and use that same definition in the report. Never blend tiers. A 96% overall rate can hide a 70% rate on your best commercial contracts. Our SLA tracking guide covers tiers and escalation.
The trap: reporting the average response time. Averages hide the misses that lose contracts.
What moves it: a dispatch decision that weighs response urgency, skills, proximity and current workload together, instead of taking the next free tech.
This isn't the same number as on-time arrival (metric 9). Response time is about the contracted promise on calls that come in unplanned. On-time arrival is about the appointment window you gave for scheduled work. A shop can be excellent at one and poor at the other, which is exactly why you track both.
4. Drive time as a share of the day
Formula: hours driving between jobs ÷ total paid field hours.
Measure it honestly: use GPS or mileage records, not what techs write on timesheets. Our guide to digital timesheets covers capturing hours at the source.
The trap: chasing the lowest possible drive time by always sending the nearest tech. That tech may lack the skill, and then you pay for the drive twice.
What moves it: matching the right tech to each job with location as one factor among several, and grouping work by service area where you can.
Why it's worth the effort: here's an illustrative example. HVAC techs earn a median $61,010 a year, about $29.33 an hour over a 2,080-hour year. BLS reports that wages and salaries make up 70% of private-industry compensation cost, with benefits the other 30%, so the loaded hourly cost is $29.33 ÷ 0.70 ≈ $41.90. Cut 30 minutes of driving per tech per day across 30 techs and 250 working days, and that's 3,750 hours, or about $157,000 a year in paid time turned back into available time.
5. Billable utilization
Formula: billable hours ÷ total paid hours.
Measure it honestly: decide once what counts as billable and write it down. Is drive time billable? Is warranty work? Is a callback? Change the definition and the trend line becomes meaningless.
The trap: pushing utilization so hard that techs rush jobs, which shows up next month as callbacks. Read utilization next to callback rate, never alone.
What moves it: fewer idle gaps between jobs, less drive time, and fewer return visits.
6. Jobs per tech per day
Formula: completed jobs ÷ tech field days.
Measure it honestly: compare within job types. A tech doing three installs isn't less productive than one doing eight tune-ups.
The trap: rewarding it on its own. Techs will skim the easy jobs, which is the cherry-picking problem that drives senior techs out.
What moves it: balanced assignment and fewer mid-day reshuffles that strand techs.
7. Revenue per tech per day
Formula: invoiced revenue ÷ tech field days.
Measure it honestly: attribute revenue to the tech who did the work, including add-on sales and agreements sold on the job.
The trap: comparing techs who work very different job mixes.
What moves it: good assignment, finished first visits, and techs who have the customer's history in hand when a recommendation makes sense.
8. Days from job complete to invoice
Formula: the average number of days between job completion and invoice sent.
Measure it honestly: measure to the invoice actually sent to the customer, not to the draft.
The trap: ignoring it because it's "an office metric." Every day of lag is cash you've earned and don't have.
What moves it: invoicing from the job record, with the parts and labor the tech logged already on it, instead of retyping a paper ticket at the end of the week.
9. On-time arrival
Formula: arrivals inside the appointment window you told the customer ÷ all scheduled appointments.
Measure it honestly: use the window the customer was actually given, not the internal target.
The trap: giving wide windows so the number looks good. Customers notice a four-hour window.
What moves it: a schedule that rebalances when the day changes, instead of one that breaks at the first cancellation.
Read your field service metrics in pairs
Every one of these can be improved by hurting another. So review them in pairs:
- Utilization with callback rate. If utilization rises while callbacks rise, your techs are rushing.
- Drive time with first-time fix. If drive time falls while first-time fix falls, you're sending the nearest tech instead of the right one.
- Jobs per day with revenue per day. If jobs rise while revenue falls, easy jobs are being prioritized.
- Response time with on-time arrival. If SLA calls get answered fast while scheduled appointments run late, emergencies are wrecking the schedule.
A pair that moves the wrong way isn't a verdict on anyone. It's a question to ask in the next ops meeting, with the jobs behind the numbers on the table.
A weekly scorecard that fits on one page
Pick five metrics, not nine:
- First-time fix
- Response time against the SLA target, by tier
- Drive time share
- Billable utilization
- Days to invoice
Show this week, last week, and the trailing 12-week average. Review the other four monthly. When one moves, ask what happened in its pair.
If you're choosing software partly on reporting, ask each vendor to show how their numbers are defined, not just how they're drawn. To size what better numbers are worth, try the ROI calculator.
Where FSM Navigator fits
Most of these field service metrics are decided at the moment a job gets assigned. FSM Navigator's intelligent dispatch evaluates SLA urgency, skills, proximity and workload on every assignment. It continuously rebalances throughout the day as conditions change, and cherry-pick prevention keeps dispatch fair across your team.
That's how response time, drive time, first-time fix and on-time arrival can move together, instead of one at another's expense. Write down your definitions first, take a baseline, and then judge any tool, ours included, by what happens to the pairs.