Busy is not a number
Jobs completed this month is the most commonly tracked figure in field service and one of the least useful. It rises when you do more small unprofitable work and falls in a month where you did three excellent installations. It tells you about motion, not direction.
The numbers worth watching all describe a conversion: work into revenue, revenue into cash, a customer into a repeat customer. Each one has a leak you can actually fix.
The seven that move margin
Start with these before anything else. Each is cheap to measure once the work is recorded properly, and each points at a specific fixable behaviour.
- First-time fix rate: the share of jobs closed on the first visit. Every point you gain removes a whole second journey, which was never billable.
- Billable utilisation: the share of a technician's paid hours that ended up on an invoice or a contract. Travel, waiting for parts and rework all show up here before they show up in the accounts.
- Average invoice value: rising usually means better quoting or captured extras; falling usually means work is being given away in the field.
- Days to invoice: hours or days between the work finishing and the invoice going out. The single most improvable number in most companies.
- Days sales outstanding: how long invoices take to get paid. Separate from the above, and often blamed for it.
- Contract renewal rate: the clearest signal of whether customers value what you do, measured before they tell you.
- Revisit rate within 30 days: how often a job comes back. High numbers point at diagnosis, parts quality or rushed scheduling, and it is worth knowing which.
Five more, once the first seven are honest
Quotation win rate, by value rather than by count, because winning many small ones is not the same as winning. Response time against whatever you promised the customer. Average job margin, which needs parts and labour costed properly to mean anything. Stock write-offs, which quietly reveal how good your van inventory discipline is. And technician retention, which is the number that determines every other number eighteen months from now.
Why most of these are wrong in most companies
A measure is only as good as the record underneath it. First-time fix rate computed from jobs somebody remembered to mark as follow-ups is a measure of memory, not of fixing. Margin computed without the parts that were bought in cash is a measure of optimism.
This is the uncomfortable part: before any of these numbers mean anything, the underlying work has to be recorded as it happens rather than reconstructed afterwards. A company that reconstructs its month at month end always reconstructs a flattering version, not through dishonesty but because the painful details are the ones that get forgotten.
A reasonable test: pick one figure on your report and try to click through to the individual jobs that produced it. If you cannot get to the records, the figure is a story rather than a measurement.
What to do with them
Pick two. Not twelve. Choose the two with the largest gap between where you are and where comparable work should be, put them somewhere the whole team sees weekly, and leave them alone for a quarter.
Metrics change behaviour, which is the point and also the risk. Push first-time fix rate hard enough on its own and technicians will stay on site too long on jobs that genuinely need a part. Pair anything you push with a counterweight you also watch.