One phone call can stall a job. One injury, one resignation, one kid home sick with a fever, and the only person who knows how to set up the seamer isn’t on the roof. If one absence can stop a crew, that’s not a crew. It’s a dependency.
TL;DR: Cross-training costs you production hours in the slow season and saves your schedule the rest of the year. When two people can do every critical task, a callout stops being an emergency and starts being a Tuesday. Build the redundancy before you need it, not during the week you lose someone.
When one man can stop the whole week
Most schedules rest on a quiet assumption that everybody shows up. Not most people. Everybody. The estimate assumed a production rate, the production rate assumed a crew, and the crew assumed one specific guy would be standing in one specific spot doing the one thing only he does.
That assumption holds until it doesn’t. Then the whole week bends around a single absence. The foreman spends his morning on the phone instead of on the deck. Somebody gets pulled off another job, which puts that job behind. The customer hears a story about weather. The rework shows up three weeks later because the fill-in guy had never done that detail and nobody had time to walk him through it.
Here’s the part that gets missed. The bill doesn’t come due on the day of the absence. It comes due later, in the crew you robbed to cover it, in the callback, in the trust you spent with a customer who’s now heard two excuses in a month. By the time you can see the cost, you can’t trace it back to the cause.
Every operation has a list of sentences that start with “he’s the only one who.” Say them out loud and you’ve got your risk register. No consultant required. The hard part is being honest about how long the list runs.
What cross training actually means in the field
Cross training doesn’t mean everybody does everything. That’s a good way to get shallow work out of a whole crew. A sheet metal mechanic who has spent fifteen years bending metal isn’t interchangeable with a laborer, and pretending otherwise insults both of them.
What it means is that every task that can stop the job has at least two people who can do it to standard. Two deep, the way a coach builds a depth chart. There’s a starter and there’s a backup, and the backup has actually taken snaps.
The useful tool here is boring. It’s a one-page matrix. Tasks down the left side, names across the top, a rating in each box. I’d keep the ratings simple:
- 0, has not been exposed to the task
- 1, has watched it done and understands the sequence
- 2, can perform it with someone alongside
- 3, can perform it alone to standard
- 4, can teach it to someone else
Fill that in honestly and the gaps announce themselves. Any row with exactly one 3 or 4 in it is a task that’s one phone call from stopping your job. Any man who’s the only 4 in three different rows is carrying weight that isn’t fair to him and isn’t safe for you.
The matrix fixes a quieter problem too. Foremen carry this in their heads, and when a foreman leaves, the map of who can do what leaves with him. Write it down and the company owns the map.
What to cross train first
Rank by consequence, not by convenience. The temptation is to train the easy things because the easy things are easy to teach. That gets you a full checklist and the same exposure you started with.
Start with tasks that gate everything else. If something has to be finished before three other trades can move, it belongs at the top of the list. Equipment setup and operation usually lands here. So does layout. So does any piece of the work with a specific sequence that, done wrong, has to be torn out instead of touched up.
Then take the tasks with exactly one holder. Difficulty has nothing to do with it. There’s often something small and unglamorous in that group, a machine nobody else has run, a piece of paperwork nobody else fills out correctly. Small tasks with one holder are the ones that surprise you, because nobody thinks of them as critical until the day they are.
Then look at anything tied to a credential, a certification, or a specific piece of documentation. Those can’t be handed off in the moment. If the qualified person is out, the work waits. Knowing that in advance is worth more than pretending it won’t come up.
What goes last is the deep craft work that takes years to build. You don’t cross train your way to a journeyman. What you can do is build a bench underneath him so the whole trade doesn’t live in one set of hands.
Where OSHA guidance fits, and where it stops
The safety side of this is worth saying plainly, because moving a worker onto an unfamiliar task is exactly when risk goes up. OSHA’s guidance on education and training notes that workers need training when they are assigned new tasks, and that peer-to-peer and on-the-job training can be effective ways to convey safe practices.
OSHA’s material on worker participation makes a related point. Workers can train their coworkers, help define safe work practices, and take part in improving the safety program itself.
That guidance covers safety training, and I want to be careful not to stretch it past where it goes. It doesn’t claim anything about your production rate, your craft quality, or whether your people stay. What it does establish is that the peer-to-peer, on-the-job method most contractors already use is a recognized way to convey safe practices, and that a new assignment is a training trigger rather than a shrug.
The practical version is mine, not OSHA’s. If you’re going to put a man on a task he hasn’t done, treat that as a training event with a name attached to it. Somebody owns the handoff. And there’s a clear line on what he’s cleared to do alone and what he isn’t.
What it does for the people doing the work
The schedule argument is the easy one to make to an owner. The people argument is what actually holds the program together, because cross training only works if the man who owns a skill is willing to hand it over.
He may not be. Some people protect what they know because what they know is why they think they’re hard to replace. Call that a character flaw if you want. It’s really a rational read of how a lot of companies operate. If a man’s only job security is that nobody else can do his job, he’d be foolish to give it away for free.
So change what you reward. Pay for teaching. Put “can teach it” at the top of the ladder instead of “can do it,” and make the pay and the title follow. The moment a 4 on the matrix is worth more than a 3, hoarding stops being smart.
The other side of it is what a wider skill set does for the man who gains it. A worker who can run three positions instead of one is worth more everywhere, and that value travels with them. Some owners hear that as a risk. I hear it as the deal being honest. People stay where they’re getting better. Nobody stays out of gratitude for being kept narrow.
Where to start this week
If you want to get moving, here’s the order I’d run it in.
- Write down every “he’s the only one who” sentence in your operation. One page. Don’t filter it, just get it out of your head and onto paper.
- Build the matrix. Tasks down, names across, 0 through 4 in the boxes. Rate honestly, including the ratings that embarrass you.
- Circle every task with one qualified person. That’s your list, ranked by how much of the job stops when he’s out.
- Pick two, not twelve. Two tasks, two backups each, this quarter. A program that tries to close every gap at once closes none of them.
- Buy the hours on purpose. Slow weeks, rain days, the tail end of a job running ahead. Put the training in the schedule as a line item, not as something that happens if there’s time. There’s never time.
- Name the teacher. One man owns the handoff and gets credit for it. Training that belongs to everybody belongs to nobody.
- Define cleared. Be explicit about what the backup can do alone and what still needs supervision. Half-trained and assumed-trained is worse than untrained, because untrained people ask questions.
- Test it before you need it. Put the backup in the seat on an ordinary day while the primary is standing right there. The first live run shouldn’t be the emergency.
The whole thing costs production hours. That’s the honest price and there’s no way around it. What you get back is a schedule that can absorb a callout, and a crew where losing a man becomes a staffing decision rather than a crisis.
You’re paying for that redundancy either way. You either pay for it in slow-season training hours, on your terms, or you pay for it in overtime, rework, and a customer who stops calling. One of those you can put on the calendar. The other one picks the date for you.
About Khary Penebaker
Khary Penebaker is Division President at MetalMaster-RoofMaster, the Upper Midwest division of Wolkow Braker Roofing Corp. He previously built Roofed Right America from startup to $35M+ in revenue with 180 employees (2014-2025) and founded Penebaker Enterprises, growing it from $1.5M to $15M. A gun violence prevention advocate and former Everytown for Gun Safety Fellow, Khary brings two decades of leadership in commercial roofing, architectural sheet metal, and civic engagement.
Common questions
What is cross-training in construction crews?
Cross-training means more than one person on a crew can do each task that matters, to the same standard. Think of it as a depth chart rather than a plan to turn everybody into a generalist. The critical tasks, the ones that stop the job when nobody can do them, need at least two qualified people behind them.
How do you cross-train field workers without losing production?
Buy the hours on purpose instead of hoping for spare time. Slow weeks, rain days, and the tail end of a job running ahead of schedule are where the training goes. Put it on the calendar as its own line item, hold it to two tasks per quarter, and name one person who owns each handoff.
What tasks should be cross-trained first?
Rank by consequence, not by how easy the task is to teach. Start with work that gates other work, like equipment setup and layout. Then take every task with exactly one qualified person, including the small unglamorous ones. Deep craft skills that take years to build come last, since you build a bench under those rather than replacing them.
How does cross-training affect employee retention?
That depends on what you reward. If a worker's only job security is that nobody else can do his job, he has little reason to teach it. Paying for the ability to teach, and making the pay and title follow, changes that math. A worker who can run three positions instead of one is worth more, and people tend to stay where they're getting better.