I built it. Nobody came.
They told us: "Build it and they will come."
I built it. Nobody came.
Twenty years in data governance. Policies. Councils. Catalogs. Frameworks with names and logos and steering committees. I was good at it. I got promoted for it. And almost none of it changed how the company actually used its data.
I'm not proud of that. But I've stopped pretending it isn't true.
The uncomfortable part
We knew. Everyone in the field knew. We'd ship a governance program, watch adoption flatline, and blame the users for not caring about data quality. So we wrote more policy. We added another review. We kept building, certain that this time they'd come.
They never did. Because we built the wrong thing. In many cases, governance was built to play defense, not to make decisions better. Those aren't the same goal. For twenty years I optimized for the first and called it the second.
AI is about to expose all of it
An agent doesn't care about your council. It needs three things: a clear rule, a clean definition, and someone who'll own the outcome when it acts. Every program that can't produce those three is about to be revealed as theater.
So here's what I'd tell my younger self. Stop building things people have to be forced to use. Build the small set of rules an agent can run tomorrow and that can keep up with the pace of change. Put your name on the result. Build that, and they actually come.
I wish I'd learned that fifteen years sooner.
What's one piece of "governance" at your company that exists to play defense, not to help anyone decide? Be honest.
Tell me. I read everything.