How Many Jobs Per Technician Per Day? Capacity Planning in 2026

TL;DR

Ask an owner how many jobs a technician does in a day and you usually get a theoretical answer: eight hours divided by an hour a job, so eight jobs. Then look at what actually happened last Tuesday and it was five, one of which ran into overtime.

As of August 2026 the practical way to answer this question is not to pick a number and schedule against it. It is to measure your real on-site duration from job status timestamps, subtract the time the day genuinely loses to travel, parts and paperwork, build the day backwards from a finish time you are willing to defend, and deliberately hold capacity back for same-day work instead of pretending you are one hundred percent booked. This guide covers each step, and the point at which better scheduling stops being the answer and hiring starts.

The theoretical number is always wrong

The arithmetic that produces the optimistic figure looks reasonable. Eight-hour day, one-hour jobs, eight jobs. Nothing in that sum is a lie; it is just missing everything that is not the job.

Here is what the same day contains in practice.

Travel between jobs. In a dense urban route this might be fifteen minutes a leg. In a rural or suburban service area it is regularly thirty or forty. Seven legs at thirty minutes is three and a half hours, which is most of a morning and an afternoon.

The morning start. Loading, checking the day, fuelling, and getting to the first job. That is rarely the zero-cost event a calendar starting at 8:00 a.m. implies.

Parts. Either a supply-house run at the start of the day, or worse, a mid-day one because a job turned out to need something not on the truck.

Write-up. Notes, photos, the invoice, the signature, the payment. Ten to fifteen minutes per job, done properly, and it is not optional if you want to get paid and defend the work later.

Callbacks and returns. Some proportion of jobs come back. Every return trip is a slot consumed by revenue you already recognised.

Add those up honestly and an eight-hour day contains four to five and a half hours of productive on-site time in most trades. That is the number your capacity plan should be built on. Everything else is scheduling into hours you do not have.

Step one: measure, do not estimate

The single highest-value thing you can do here is stop guessing your average job duration.

Estimated durations are systematically optimistic, and the reason is human rather than technical: people remember the job that went cleanly and forget the one where the access was bad, the customer wanted to talk, and the part was wrong. Ask a technician how long a standard service takes and you will get the time it takes when nothing goes wrong, which is not the average.

The data is already sitting in your system. A job status lifecycle that moves through Unassigned, Dispatched, En Route, Arrived, In Progress and Completed produces a timestamped record of every job, and the gap between arrived and completed is your real on-site duration. Pull three months of completed jobs, split by job type, and look at the distribution rather than just the mean.

Two things to look at specifically:

• The median, not just the average. One catastrophic six-hour job drags the average somewhere unhelpful. The median tells you what a normal one looks like.
• The spread. A job type that runs 45 to 75 minutes is schedulable. A job type that runs 30 minutes to four hours is not one job type, it is two or three that share a name, and splitting them out will improve your scheduling more than any other single change.

Do the same for travel. The difference between planned and actual arrival times, or a trip history from vehicle tracking, gives you a real average leg rather than an assumed one.

Step two: build the day backwards

Most schedules are built forwards. Start at 8:00, add a job, add another, keep going until the day looks full. That method has no natural stopping point, which is precisely why it overfills.

Build backwards instead.

1. Pick the finish time you are willing to defend. Not the time your technicians sometimes finish. The time you are prepared to have as normal. Say 4:30 p.m. for a 7:30 a.m. start.
2. Subtract the fixed costs of the day. Morning load and first travel leg. Lunch, which is a legal and human requirement, not a buffer to be raided. Any end-of-day return or restock.
3. Subtract travel between the jobs you plan to run. Legs equal to jobs minus one, at your measured average.
4. Divide what is left by your measured on-site average, plus write-up time.
5. Subtract the slack you are holding for same-day work.

What comes out is a smaller number than the forwards method produced, and it is the honest one. Working through it with realistic figures, an eight-hour day at forty minutes travel and ninety minutes on site plus fifteen minutes of write-up does not hold five jobs. It holds three, possibly four if the day is geographically tight.

That is not a failure of the business. It is what the day actually contains, and knowing it is what lets you price correctly and promise accurately.

The mechanical version of step two — putting lunch, admin and drive time on the calendar as real blocks rather than assumptions — is covered in blocking lunch, admin and drive time. It is the change that makes the honest number visible to whoever is booking.

Step three: attack travel before you attack anything else

Once you can see the real split, travel is usually the biggest single non-productive block in the day, and it is also the most improvable.

Two levers do most of the work.

Geographic clustering. Booking by the order the phone rang produces a day that crosses the service area repeatedly. Booking by area produces a day where the legs are short. The same jobs, the same technician, a materially different day. This is the entire premise of route optimization, and in a suburban service area it commonly converts an hour or more of driving into another job.

Knowing where the trucks actually are. A dispatcher deciding who takes an incoming job by looking at empty calendar slots is answering the wrong question. The right question is who can be there soonest, which is a function of position, not availability. Live technician mapping and traffic-aware ETAs turn that into a fact rather than a guess, and they make the arrival window you give the customer defensible. GPS tracking pays for itself in this specific decision more than in any other.

If you fix nothing else, fix travel. It is the cheapest capacity you will ever buy, because you already own it.

Step four: hold capacity back on purpose

This is the part that feels wrong to owners and is almost always right.

If same-day and emergency work is a real part of your revenue — and in most trades it is the highest-margin part — then booking every technician to capacity is a decision to handle every emergency badly. The call comes in, the day is full, and the dispatcher either turns down profitable work or reshuffles three customers to fit it in.

The alternative is a deliberate hold. One slot per technician per day, or a half-day across the team, left unbooked in the schedule.

Two objections, both answerable:

"I am giving away capacity." Only if it goes unused. A held slot still open at 3 p.m. the day before can be filled with flexible work — a maintenance visit, a deferred callback, a quote appointment. You have lost nothing except the ability to book it three weeks out.

"We will just fit emergencies in." You will, and the cost is invisible. It lands on the customers who get moved, the technician finishing after dark, and the arrival windows that stop being accurate. The framework for deciding what actually outranks what is in emergency versus scheduled jobs.

The general labour-market context here is not obscure — occupational data from the Bureau of Labor Statistics is the standard reference for how skilled trades staffing is trending — and every owner already knows the practical version: technicians are hard to find and harder to keep, and burning the ones you have to protect a booking number is a poor trade.

A booked day and a productive day

It is worth being precise about the difference, because the two get used interchangeably and they describe different things.

The reason overbooking survives is that it looks like progress. The schedule is full, the day looks productive, and the costs of the gap between the two columns land somewhere that is not the calendar: in unpaid hours, in rushed work that comes back, and in customers who stop believing your arrival windows and start not being home for them.

That last one is worth naming plainly. Chronic lateness produces no-shows. A customer who has been told 1 p.m. twice and seen 4 p.m. twice will run an errand during the third window. The no-show is then recorded as the customer's failure when it was manufactured by the schedule.

Utilisation is a target, not a goal

There is a temptation to reduce all of this to a single utilisation percentage and drive it upward. Be careful with that.

Utilisation measured as billable on-site hours over paid hours is a genuinely useful diagnostic. It tells you where the day goes. It is a poor target, because the fastest ways to improve it are all bad: cut write-up time so documentation degrades, cut travel by refusing outlying jobs you would want, or cut the buffer that absorbs same-day work.

Use it to find the problem, then fix the process. A team at a modest utilisation with short legs and clean documentation is in better shape than one at a high utilisation held together by unpaid evening hours.

When the answer is hiring

At some point better scheduling stops returning anything and you need another person. Three conditions, all of them, before you hire:

Realistic capacity is genuinely full for several consecutive weeks. Not one busy fortnight, and not the theoretical capacity — the honest number from step two.

Travel is already reasonable. If technicians still spend large blocks of the day driving or collecting parts, routing and process work will buy you more capacity faster and cheaper than a hire, and it will buy it in weeks rather than months.

You are declining or delaying work you would want. Turning away out-of-area jobs is fine. A two-week lead time on core work is a capacity signal.

When those hold, the arithmetic is straightforward: your measured jobs-per-day times your average job value times working days, against fully loaded cost. Just run it with the honest capacity figure, because running it with the theoretical one is how a new hire ends up looking underproductive when the schedule was never realistic to begin with.

A useful intermediate step is a dispatcher, or better dispatch tooling, before a technician. One person deciding assignments off a live map is often worth a meaningful fraction of a technician in recovered capacity — the options are surveyed in best dispatch software for service companies.

What this costs

GetTimePad is $79/mo Starter for one staff member, $199/mo Pro for up to five staff — which adds live GPS tracking and traffic-aware ETAs, payments, review routing, advanced automations and the AI receptionist — and $499/mo Agency for unlimited staff with multi-location and API access. Extra seats are $25/mo on Starter and Pro, the SMS Bundle is $19/mo, the GPS device is $299 one-time with $15 per device per month connectivity, and annual billing gives you two months free. There is a 14-day free trial and a live demo with no signup required. Full details are on pricing.

For capacity work specifically, the Pro tier is the relevant one: the job status timestamps that give you real durations, the live map that makes dispatch decisions factual, and the dispatch board that shows who genuinely has room.

Where to start

Pull three months of completed jobs and calculate the median arrived-to-completed time by job type. That number alone will change how you book.

Then put travel, lunch and admin on the calendar as real blocks and see what the day actually holds. Then set one held slot per technician for same-day work and leave it there for a month. Then look at whether your finish times moved, whether overtime dropped, and whether arrival windows started holding. Those three are what capacity planning is actually for.

Frequently asked questions

How many jobs should a technician do per day?

There is no universal number, and any figure quoted without the trade, the average job length and the density of the service area attached to it is guesswork. The useful method is to divide the productive hours you actually have, after drive time and administration, by your measured average on-site duration, and then hold back a slot for same-day work. Most small field-service teams find the honest answer is one to three jobs lower than the theoretical one.

How do I measure real average job duration?

Read it out of your job status timestamps rather than estimating it, using the gap between arrived and completed across the last few months of finished jobs of the same type. Estimated durations are optimistic almost universally, because people remember the job that went smoothly and forget the one where the part was wrong, and a schedule built on an optimistic average is overbooked before the day begins.

What is the difference between a booked day and a productive day?

A booked day is one where every slot on the calendar has a job in it. A productive day is one where the jobs in those slots were actually completed, on time, without overtime or a return trip. The two diverge whenever the schedule ignores travel, parts collection, write-up and callbacks, and the gap shows up as late arrivals and unpaid hours rather than as anything visible on the calendar.

How much slack should I leave for emergency and same-day work?

If same-day work is a meaningful share of your revenue, leave a deliberate hold in each technician day rather than booking to capacity and reshuffling when the call comes. A held slot that goes unused can be filled the afternoon before with flexible work, whereas a fully booked day forced to absorb an emergency costs you three displaced customers and a technician finishing after dark.

Does overbooking technicians actually increase revenue?

Rarely, and usually only in the first week. Overbooking converts into late arrivals, unpaid overtime, rushed work that produces callbacks, and customer no-shows caused by arrival windows nobody trusts any more, all of which are costs that do not appear on the schedule. A calendar that reflects reality produces fewer booked jobs and more completed ones.

When should I hire another technician instead of scheduling better?

Hire when your realistic capacity is genuinely full for several consecutive weeks, drive time is already reasonable, and you are turning away or delaying work you would want. If technicians are spending large parts of the day travelling or collecting parts, routing and process changes will usually buy you more capacity than a new hire, and they buy it in weeks rather than months.