Mud & Microchips · July 2026

Will My Crew Need AI Training — And How Much of My Time Will That Cost?

The first question is about your crew. Will they need training? Yes — but probably not the kind you're dreading, and less of it than you think.

The second one is about you. How much of your time will it cost? If you do it right, almost none — and that's the whole point. If this project eats your life for a quarter, it was set up wrong from the get-go.

Let's dive into this.

Start Before the Tool: The Conversation You Can't Skip

Here's the mistake we see over and over — and you most likely experienced it long before you ever heard about AI. It happened with your software, and it'll happen with AI too if you let it.

An owner or a manager gets excited about a new tool. They rip out the old process, drop the new app in front of the team, and say some version of "here's the thing, just use it." Then they walk away and assume it's handled.

It isn't. The second you leave the room, people go back to the way they've always done it. And it's because nobody gave them a reason to change, and the old way still works well enough to keep doing it.

This isn't just our opinion from the jobsite. Workplace research and Brené Brown's work on leadership both land in the same place: you don't get people through a big change by giving orders and expecting compliance. You get them through it by bringing them along. People commit to what they help build. They resist what gets done to them.

So before you train anybody on anything, you have a conversation. And we'd actually split it into two.

The First Conversation: Candor

This one is honest and a little uncomfortable, as it should be.

We recently sat down with two owners who told us that five years into their business, they still don't see whether a project was profitable until it's finished. Five years. That's the kind of thing you can only fix if the people doing the work understand why it matters.

So you tell your foremen the truth about where the company actually is. Here's what's working, here's what's costing us, and here's the thing I can't see clearly that's hurting us.

What you don't do is turn it into a threat. No scarcity, no tension, no "if we don't do this we'll have to close and you'll lose your job." That's manipulation, and your crew will smell it instantly. It might get you compliance for a week. It won't get you buy-in for any new changes you want to introduce.

The goal is transparency and visibility: giving people enough of the bigger picture that they can make their own informed decisions. You want them to feel like they chose this — not like they were marched into it.

The Second Conversation: Bring Them In

The first conversation is you being honest. The second one is you shutting up and listening.

This is the brainstorm. You put the problem on the table and you let the foremen contribute. And here's what surprises a lot of owners: the people doing the work every day very often have good ideas about how to fix their own systems. They've been living inside the friction. They know where the wasted steps are. Nobody's ever asked them.

When a foreman helps shape the solution, he's not being told to use a new app. He's using the thing he helped design. That's a completely different relationship to the change — and it's the difference between adoption that works and (another) tool that gets abandoned.

The Training Itself Is the Easy Part

Now to the part everyone worries about, and it might be disappointing: the actual training is a few hours.

That's it. AI is practical. Teaching your crew to use it is more like teaching them to use a new power tool than sending them back to school. You show them what it does, they use it, they see the result. People get on board when they see it work.

The real work is what comes after.

The Feedback Loop: Don't Pocket the Time Savings Yet

Say a tool takes a job that used to have ten steps — imagine pulling data from ten different places to build one report — and cuts it down to two. Your instinct is to grab those saved hours immediately and move on. Don't. Not yet.

For the first three months, have your person spend the same amount of time on that task as before. Except now they're going over what the tool produced, checking it against what it should be, and writing down anything that's off. Everything that's not "quite right."

That feedback goes back to your provider, and the tool gets reiterated and tuned to how your business actually works. Dedicate one person to this. Keep them on it for three months.

Then the time starts to shrink. As the tool gets better, what took an hour drops to half that. Eventually it's ten minutes tops, with your person doing a final approval at the end. That's where you finally pocket the savings.

Where This Actually Costs You — the Owner

Here's a real example: a five-person shop, providing services into construction. After our initial conversations and the offer — no more than two hours of the owner's time — the owner said four words: "Do it. Don't involve me."

And that was correct. The training, the feedback loop, the reiterations — all of it got handled by a dedicated person on his team, and all of it got documented for whoever joins the company down the road. Full visibility. He didn't need to be in the room for any of it.

Because here's the hard truth, and we'll say it as a peer who's watched a lot of good operators get stuck in this exact spot: if you can't delegate this, the problem isn't the AI. It's that you've built yourself a job instead of a business. The whole value of a rollout like this is that it doesn't run through you. Two hours up front to make the call and set the direction. Then you hand it off.

If you're the bottleneck on your own AI implementation, you've recreated the exact problem you were trying to solve.

Document It, or Watch the Knowledge Walk Out the Door

You'll notice "properly documented" keeps coming up. It's SOPs — standard operating procedures — and done right, they cost you almost nothing to create.

The trick is to build them parallel to the work, not as a separate project. As your person learns the tool and runs the process, they either write down the steps as they go or record a quick video of themselves doing it. Then you have someone — an office manager, for example — make sure those files are stored properly and named so they can actually be found later.

When all your business knowledge lives in people's heads, there's a risk. If that person leaves, the knowledge leaves with them. And it shows up when you want to sell.

We've seen it. A business gets evaluated for pennies compared to what it's actually worth, because everything that makes it run is locked inside a few employees' heads. Documentation is how you make sure the value you built is value someone can actually buy.

The Honest Answer, Straight to Your Face

If you cornered us on a jobsite and said "just tell me honestly, is this going to be a nightmare," here's what we'd say.

The only one who likes change is a wet baby. It's not going to be easy. This is one of those projects where you have to trust the process for a while before you get your good night's sleep back. Expect a lot of learning and unlearning: forgetting the old ways, keeping your mind open, letting people find their footing.

Is it worth it? If you look at your business on a ten, fifteen, twenty-year horizon: yes.

Now, the thing to stop worrying about, and the thing you actually should.

Stop worrying that this will eat your time. It won't, if you delegate it the way you should delegate anything else.

Stop worrying about people leaving over it, too. The ones who leave were usually the ones already halfway out the door.

What you should prepare for is the human side: the buy-in, the honest conversations, the patience while your crew comes around. Get that part right and the rest follows. Get it wrong and the best tool in the world sits unused, like that software you're still a little sore about.

And the good news, again: AI is practical. It's a power tool, not a philosophy. Your crew will see the results and come around in their own time. They almost always do.

If you want the numbers on exactly how much labour time a rollout like this can actually give you back, we broke that down in our post on whether a small construction firm can afford AI. This piece is about the people. That one is about the math.

Mud & Microchips helps construction and trades businesses implement AI that actually works — starting with an operations audit that measures where your hours actually go.