How You Know You Have a Successful AI Champions Program.

We just closed a six-month engagement with about thirty AI champions at a client. Every other week for six months, we showed up and ran roughly the same meeting. We talked through what actually happened in AI since we last met and whether it mattered to them or not. We reviewed our metrics as a group, so everyone could see whether we were on track for what we set out to do together. We taught something, usually a feature or a technique that would make the tools more useful in their real work. And we left time for show and tell, where people showed what they had been building.

The champion, what we call an operator, is the front-line driver of AI discovery and adoption inside their own function. They are the ones who surface the use cases their department actually has, the pipeline our team then helps turn into level one, level two, and level three solutions. We build real tools in that process, and some of them genuinely change how much a team can get done. But the tools were never the point.

The point was two things. First, opening people up to what's possible when they bring these tools into their work, and building the confidence to go do it. Second, and this is the one that matters more, growing their own capability until they're the ones driving it, surfacing their own ideas and building their own solutions.

You find out whether that worked at the last meeting.

The last meeting is one slide

We end every program the same way. We ask each person to build a single slide with four boxes: where we started, what changed, where we are now, and what's next. That's the whole prompt. We don't tell them what to put in the boxes. We don't ask them to mention anything we built. We hand them the quadrant and we sit back.

I've run this enough times now to know what comes out, and it still gets me every time.

People barely talk about the solutions. They mention them, sure. But what they choose to spend their slide on is how they changed. Nobody prompts this. It happens over and over, across every version of this program I've run.

What comes out is a personal story. Someone starts with AI as a thing they never really used, or used casually as a slightly better search engine. By the "where we are now" box, it's a part of how they work that they'd have a hard time giving up. They walk you through the middle themselves: the first time it helped them make something better than they could have made alone, the point where they started doing more and holding a higher bar for the work, the week it went from novelty to how they work. You don't have to draw it out of them. The blank quadrant does it.

That's the read I trust most. Not the hours saved, not the list of tools, though those are real and we track them. When you give a person a blank slide at the end of six months and they use it to tell you they're different, you know it took.

What I told them at the close

I closed the final meeting with a few hopes, and I mean the word.

I hope you went on a journey. I hope you learned a few things. I hope you got comfortable taking chances and exploring, and that you see how much the human in the loop still matters. I hope you're enjoying the work in a new way, and creating more than you used to, because the ideas in your head have a faster path out now. I hope you feel good about where you stand in all of this.

Then the part I most wanted them to keep. The tools are going to change. What you built these six months is the muscle for learning how to use whatever comes next, and that muscle doesn't expire when the next model ships. The building blocks stay. Knowing what a context window is and how to fill it well. Learning to give the tool the context it actually needs to help you. Spotting a hallucination when you see one, and being willing to push back instead of taking the first answer. Understanding what the models are, how the features across these different applications work, and that the technology isn't perfect and won't be. And knowing that you only get good at any of it by using it. Those transfer. Every tool you'll touch for the rest of your career sits on top of them.

The largest list of solutions this company had ever surfaced came out of this group. We were told as much. That belongs to them, not to us. We just showed up every other week and held the space for it.

Why we run it this way

If we measured ourselves only by the tools we build, we'd be optimizing for the wrong thing. A tool solves one problem. A person who has learned to solve for themselves solves the next hundred. This is the whole point. If we do our transformation job well, we work ourselves out of a job. That trade is the one we're always making inside these programs, and it's why the quadrant at the end matters more to me than any single number on the metrics slide.

The champions are at the beginning, not the end. The program closes and the journey keeps going. They know how to find the use cases in their own work now, how to reach for a tool without fear, how to teach the person sitting next to them. That's what we were building the whole time. The solutions were just what we made along the way.