The Owner Who Trains Agents: A New Legacy in the Trades

Ask a trades owner what legacy means and you'll hear a lot of familiar answers. The trucks with your name on the side. The building. The reputation in town. The business that gets handed to the next generation.

All of that still matters. But there's another piece of the business that has always been much harder to pass down: what the owner knows. How you make the call when the answer isn't obvious. How you handle an unhappy customer. What you look for before approving a job. The standards you refuse to compromise. The hundred little decisions you make because you've spent 20 or 30 years learning what works.

For most of the history of the trades, transferring that knowledge meant training another person and hoping enough of it stuck. AI agents introduce another possibility.

In a recent Just Start AI training series, we built AI agents in two different environments, starting with a personal assistant in Muse and then moving into more advanced agents with Grok. The session was about building agents, but underneath the technology was a much bigger idea: What happens when the knowledge you've spent your career building becomes something you can actively teach your business?

Watch the Grok session: https://youtu.be/H0tnR_m3TaQ?si=J7Jt0vqnaAn0sSSJ

What Owners Have Always Actually Built

Strip away the equipment and real estate, and a great trades business has always been built on more than physical assets. It's customer relationships. It's reputation. And it's the way the company works.

That last piece can be difficult to define because so much of it becomes instinctive over time. An experienced owner doesn't necessarily pull out a manual before making a decision. They know what to look for. They've seen the problem before. They recognize when something doesn't feel right.

The same is true throughout the company. A great service manager develops judgment. So does an experienced dispatcher, comfort advisor, installer, CSR or technician. Some of that knowledge gets documented, but a lot of it doesn't.

That's where agents get interesting.

An AI agent trained on your business creates another way to capture parts of what used to live almost entirely in people's heads: how decisions get made, how customers are handled, what standards matter, what information should be checked and what happens next. That's more than automating a task. It's a new way to preserve how the business works.

Your Judgment Doesn't Scale Unless You Teach It

During the session, Michael Parker showed an agent he'd been training to review designs. He didn't simply tell it, "Review this design." He performed reviews in front of it, showed it what he looks for and continued testing its feedback. Over time, the agent became capable of identifying many of the same things he would identify himself.

The goal wasn't to make Michael irrelevant. It was to make his judgment available in more places.

That's leverage.

An owner already does this with people. You train a manager so they can make decisions without calling you every five minutes. That manager trains someone else. Eventually, one person's standards can influence dozens or hundreds of decisions they never personally touch.

Agents create another layer of that leverage, but they don't magically know how your company works. You have to teach them. That means defining the job, providing context, showing them what good and bad look like, setting boundaries and continuing to refine the way they operate.

The Shop That Outlives Its Founder

Succession has always been one of the harder problems in the trades. When an experienced owner steps away, the business doesn't just lose a person. It can lose decades of judgment that was never fully documented.

Imagine an owner spending the final years before an exit not only running the company, but deliberately training it. How do we price this? When does something need to be escalated? What does a great customer experience look like? What information needs to be checked before making this decision? How do we follow up when an estimate doesn't close? What do we do when the normal process doesn't fit the situation?

Some of those answers belong in SOPs. Some belong in employee training. Some need to remain human decisions. But increasingly, some can also become instructions, examples, rules, workflows and context given to agents that help the next person operate the business.

The successor doesn't have to inherit a filing cabinet full of procedures nobody opens. They can inherit systems that help put those procedures to work, alongside the people who understand the company and continue developing them.

The name on the truck still matters. The reputation still matters. The people absolutely still matter. But the intelligence inside the business is part of what makes all of those things valuable.

Legacy Has Always Been About Leverage

A business already represents leverage. One person's standards become the standards of a crew. A crew becomes a company. The company serves thousands of customers. Eventually, the founder can step away from individual jobs because other people know what good looks like.

AI doesn't change that principle. It changes how far it can extend.

In the session, the examples started small: an agent that creates a daily report, one that reviews work, one that pulls information together and one that follows a defined process. That's also where owners should start, rather than asking how to automate the entire company.

Start with something more specific: What do I repeatedly have to look at, decide, explain or correct?

Then ask what someone would need to know to handle that correctly without you. What information would they need? What rules would they follow? What does a good result look like? What would require escalation? What mistakes have you learned not to make?

That's the knowledge worth capturing. Once you can explain how you think about the work, you can start teaching it.

Your Homework This Week

Write down three standards that define your company. Not generic values like "quality" or "customer service." Get specific about what those ideas actually mean inside the business.

How should a lead be answered? How should a complaint be handled? What should happen after a job is completed? What does your best manager know to check that a brand-new manager wouldn't? What do you notice immediately that someone with five fewer years of experience might miss?

Pick one and try to document it well enough that someone else could make the decision without asking you.

That document could become an SOP. It could become training for your next employee. Increasingly, it could also become part of the blueprint for an agent that helps your team carry out that standard consistently.

The bigger question is what your business knows today that only you know: how you make the call, how you handle the customer, what you notice that someone else misses and what "good" looks like before anyone has to ask.

Start there. The next generation of leverage isn't just about getting more work done. It's about teaching the business what you know.

Watch the full session: https://youtu.be/H0tnR_m3TaQ?si=J7Jt0vqnaAn0sSSJ