Knowing how much work you can take

It is Tuesday and four jobs are already on the board for the week. A homeowner calls about a $1,900 panel swap and wants it Friday. Friday is blank on the calendar, so you say yes.

Friday was not open. Friday was where Wednesday's job was going to spill, because Wednesday's job always spills. You find that out Thursday afternoon, and then you have to call the homeowner.

Capacity is not the blank space on your calendar. It is the hours you can put a tool in your hand and still finish what you already promised.

Where a 45-hour week actually goes

Start with an honest count. Below is one worked example: a one-person van trade, seven to four, five days a week. The numbers show the method, not an average.

Twenty-six billable hours out of forty-five. That ratio is not a failing. It moves with the work rather than with how hard you push: a crew parked on one site for two weeks loses almost nothing to driving, while a service business sent to a new address every two hours runs far below 26. Write down where one ordinary week actually went and you have your own number.

The number that matters for scheduling is one step above billable: field hours — site time plus the driving and material runs around it. Here that is 34.5. Office and estimating time gets done at night or Saturday morning whether you plan it or not, so leave it out of the day rather than pretend it fits between jobs.

How many jobs of your size fit in what's left

Run every job type you sell through one calculation: on-site hours plus its share of drive and material time. That total is a slot.

Small jobs carry about an hour of overhead each. Eleven service calls carry eleven hours of driving and material runs; one week-long remodel carries under four. That is 22 productive hours against 32, for the same days on the clock — usually priced as if they were the same.

Mixing sizes wastes the remainder. Put a 5.75-hour job in a nine-hour day and the leftover three hours are not a job, they are a gap. The most profitable weeks are usually the boring ones: same job type, same size, back to back.

Book 80 percent, and measure your own 20

Book 80 to 85 percent of your field hours. That is a common rule of thumb rather than a researched figure; on 34.5 field hours it comes to 28 or 29 bookable hours a week. The rest is not idle time. It absorbs the thing that goes wrong, and something goes wrong most weeks — a part comes in wrong, an inspector reschedules, a wall opens up worse than the walkthrough suggested.

You do not have to take 20 percent on faith. Go through your last ten completed jobs and write the field hours you estimated next to the hours it took. Add both columns and divide. Ten jobs at 128 hours against 108 estimated means you run 19 percent over, and that is your buffer — measured, not borrowed from an article.

Callbacks are the biggest single draw on that buffer. A callback on a four-hour job is rarely a four-hour problem: the drive out, an hour or two of work, sometimes a part, the drive back. Call it three field hours. One a week is nine percent of 34.5. Two is a whole job slot.

Overbooking also compounds. A crew working past six to catch up makes more of the mistakes that generate callbacks, which eat the capacity you were trying to create. Booking 40 hours instead of 28 gives you the same work delivered late, and sometimes a smaller invoice than the week before.

Measure your backlog in weeks, not in jobs

"I've got a lot on" is not a number. This is:

Backlog (weeks) = field hours of signed, unstarted work ÷ bookable field hours per week

Ninety-six hours of signed work against 28 bookable hours is 3.4 weeks. Recalculate every Friday and you have a trend instead of a feeling.

What counts as healthy depends on your trade and your market. Under a week and you are exposed: one cancellation and you are calling old customers. A few weeks gives you room to plan material and turn down bad jobs while customers will still wait. Too long and you lose work you never hear about, because a homeowner with a leak calls the second name on the list. Emergency work runs out of road faster than remodeling does, so pick your own ceiling and watch your close rate as you approach it.

Seasonality shows up in the same data. Pull last year's invoices by month. If April was double February, this year's April was decided in February — that is when to raise the price or line up help, not in May.

Slammed for three weeks is weather. Slammed for six months is a decision.

A short spike you ride out. A backlog that has sat above your ceiling since spring means demand exceeds capacity structurally, and you have four responses with very different math.

Raise the price. While demand keeps outrunning your calendar, this is the cheapest of the four: the leads you lose are ones you could not have served anyway. Run it on your own rate. If 28 bookable hours at $95 is $2,660 a week, going to $110 and losing a fifth of your leads still fills 28 hours — now worth $3,080. (Arithmetic on an example rate, not a going rate; hourly rates swing widely by trade and region.) You did not work more. You stopped rationing by first-come-first-served. Work already under contract stays at the price you signed.

Give a real date instead of a vague one. "About six weeks" loses the job. "I can start the week of October 12" holds it. Keep deferred jobs in a written list with the date you promised.

Send the overflow to someone you trust. You lose the job and keep the customer.

Add capacity, and do it last. A new hire does not add 40 hours. Expect well under half of that at first, while your own billable time drops as you supervise, so net capacity can go backwards before it goes forwards. Payroll tax deposits, state unemployment registration, new-hire reporting and — in most states — workers' compensation all start with the first paycheck. Comp rules vary most from state to state, so check your state's agency and your insurance agent before the start date, not after.

Putting a date you can hit on the estimate

You make a scheduling promise in the estimate whether you write it down or not. If your backlog says three and a half weeks and the estimate says nothing, the customer assumes next week. Put the start window and your assumptions about site access and material lead time in the estimate itself in DoneQuote, next to the price, so the date you worked out from your field hours is the date the customer approves.

Being fully booked is not a problem to solve. It is a signal that your price, your job mix or your capacity is out of line with your demand. Knowing which one is the whole job.