My Way or the Highway?

This week, the company I have spent years inside became a different company. Same buildings, same people, a new name over the door.

I am stepping away from it soon, and stepping away clears your eyes. It takes away the thing you were selling. No method to defend, no budget to protect, no case left to make for the way I did it. So I will say what I believe, now that saying it costs me nothing.

A few weeks ago, at a conference, I watched someone from another company stand on a stage and sell the one true way to transform an enterprise with AI. He was good. He was certain. And he was selling the opposite of everything we built here.

There are two camps in this fight. One says put a small team at the center and change the culture from the inside out. The other says keep the function outside the org and let it spread team by team, winning one department before it moves to the next, changing people while they keep the lights on. We built the first. The man on the stage was preaching the second, as gospel.

He was right about his way. He was wrong that it is the only one.

I ran the inside model for three years and it worked. It also hit walls his model would have walked straight through. And his model has failures ours would have caught. Neither of us found the one true way, because there isn’t one. There is only the way that fits the company you actually have.

That is the part nobody selling from a stage will tell you. What makes or breaks a transformation is not the method you pick. It is where your organization fights change, and whether your method meets it there.

If your leadership will not commit, no amount of grassroots enablement saves you. If the middle is frozen, a mandate from the top dies on the way down. If your people are willing but unequipped, the best strategy in the world means nothing until you teach them.

Find the fight first. Then choose, and be willing to run more than one play.

This is true if you are a company of a hundred thousand. It is just as true if you are a company of one.

Figuring this out is the hardest part of AI, and change management is right beside it. Really, they are one problem, because one precipitates the other. The method sets the change in motion, and the change decides whether the method survives.

We built one version of this. It worked, and it was never the only way it could have. Whatever you are running, stop looking for the gospel.

Go find the way that fits what you have.