Dispatching and Operations
Crew Scheduling for roofing contractors
Four crews and four jobs only balances if any crew can take any job. In roofing, none of them can.
A week that can actually be run
Work matched to the crews able to do it, not distributed evenly across the ones who exist. Type of roof, the kit that has to be on site, any ticket the job requires, and the practical question of who has done this before.
Capacity counted before anything is promised to a customer. The moment a date is given out, the schedule stops being a planning document and becomes a set of commitments, and reversing one of those is expensive in a way that moving a line on a plan is not.
And one version of the week, issued rather than circulated. Where three copies exist, the crew is working from whichever one they were sent last, and nobody in the office knows which that was.
Capacity is a total; the week is a fit
Ask an owner how much they can take on and the answer arrives as a number of crews. It is the natural unit and it is the wrong one, because it treats the resource as fungible. Roofing resources are not fungible in several directions at once, and each one narrows the set of crews that can take a given job.
Some of it is skill: a team that does tile all week is not the team you want learning it on a customer’s roof. Some of it is equipment, which travels with a particular truck and cannot be in two places. Some of it is licensing or a required ticket. Some of it is simply that one lead can hold a conversation the homeowner will understand.
The consequence is that a schedule which looks comfortably loaded can be impossible, and the impossibility is invisible until somebody tries to run it. The constraint was never the number of crews — it was that the two jobs needing the same capability landed in the same week, and the planning tool had no way to know they were competing.
So the work here is unglamorous and specific: write down what each crew can genuinely do, keep it current, and schedule against that rather than against a count of crews. It is also the reason this desk produces a list of where the match keeps failing — repeated conflicts on one capability are a hiring or training answer, not a scheduling one.
The distance between two jobs is paid time
One input to a schedule is routinely treated as free and is not. Sending a crew from one site to another during the day is compensable time, and the rule is explicit about it. Under 29 CFR 785.38, “Time spent by an employee in travel as part of his principal activity, such as travel from job site to job site during the workday, must be counted as hours worked.”
That changes how a week should be assembled. Two jobs an hour apart and two jobs on the same street are not equivalent pieces of work even where the roofs are identical, and a schedule built purely on job duration has quietly bought the difference without pricing it.
The same logic reaches further out. At 29 CFR 785.39, “Travel away from home is clearly worktime when it cuts across the employee’s workday. The employee is simply substituting travel for other duties.” Storm work and out-of-area jobs are where this bites hardest, and they are exactly the jobs scheduled on distance alone because the work itself looked worth the trip.
Both provisions bind the employer, which is you rather than this brand — nothing on this desk employs your crews, sets their pay or determines what any individual is owed. They are here because they change what a good schedule looks like: routing is part of the plan, not an afterthought, and how any of this applies to your arrangements is a question for your own adviser.
This describes two federal wage-and-hour provisions in general terms. It is not legal advice, it makes no determination about your employees, your pay practices or your obligations, and other rules may reach the same hours. Those questions belong with your own counsel.
The plan the rest of the day answers to
This desk sits inside Dispatching and Operations, and the boundary that governs the whole department — including the point at which coordinating work would become directing how it is performed — is set out there rather than repeated here. What is particular to this page is that it happens before the week starts. Everything that follows is a departure from what was decided here.
What owners ask about the week ahead
We have enough crews. Why does the week still fall apart?
Because having enough is a statement about totals and the week is decided by fit. Four crews and four jobs balances only if any crew can take any job, and in roofing they cannot — steep and tile, low-slope, anything needing a particular ticket or a particular truck. The schedule that balances on a spreadsheet is often the one that cannot be run on Monday.
What do you need from us to build one?
What each crew is actually able to take, honestly rather than aspirationally, and what equipment travels with which team. Then the work, with its type and its constraints. Most companies have never written the first part down, and the exercise of doing it once usually explains several past weeks.
Do you decide who is on which crew?
No. How crews are made up is a decision about people, and it belongs with you and whoever runs production. This desk takes the crews as they are and matches work to them, then tells you where the match keeps failing — which is a useful input to your decision without being a substitute for it.
Our schedule changes constantly. Is scheduling even worth doing?
It is worth doing precisely because it changes. A week nobody planned cannot be replanned; there is no baseline to move things against and no way to see what a change costs. The value is not that the plan survives — it is that when the plan breaks, somebody can tell within a minute what else has to move.
What about the travelling between jobs?
It is scheduled explicitly rather than absorbed, because it is real time with real cost attached, and there are federal rules about how it counts. Two jobs an hour apart is not the same as two jobs on one street even where the work is identical. The section below sets out the rule we work to.
How far ahead should the schedule go?
Far enough that material and permits can land in front of it, which in practice means further than most companies plan and less far than they think they should. A schedule extending months out is a forecast wearing a schedule’s clothes. What matters more is that the near weeks are firm and everybody is reading the same version of them.
The travel-time provisions quoted here
- 29 CFR 785.38 — Travel that is all in the day's work — U.S. Department of Labor, via eCFR
- 29 CFR 785.39 — Travel away from home community — U.S. Department of Labor, via eCFR
List what each of your crews cannot take
If that list has never been written down, it is the reason the same week keeps going wrong in the same way.