There’s a moment in almost every architecture review where someone puts up a slide with a Bronze, Silver and Gold layers, or a tiering model with tidy boxes labelled Data Products, and I watch the room nod along without actually agreeing to anything. The numbers are right. The logic holds. And still, nothing moves.
I used to think that was a communication problem. Explain it better, add a diagram, simplify the language. I’ve stopped believing that.
What I Thought Would Convince People
For years my instinct in these conversations was the same: get the model right, and the model will do the convincing. If the medallion layers are rationalised cleanly, if the tiering makes sense on paper, people will follow the logic to the same conclusion I had. That’s how I was trained to think, and honestly, it’s how most of us in this line of work are trained to think.
Analysis first. Persuasion is just analysis, presented well.
I was reading Steven Johnson’s Farsighted a few weeks ago, not thinking about work at all, and one line stopped me. He argues that the hard part of a decision isn’t the mapping or the modelling, we’re pretty decent at those. It’s the part after. The part where you have to actually feel your way into a future you haven’t lived yet. And for that, he says, novels do something spreadsheets can’t. They let you experience a choice from the inside.
The Room That Didn’t Move
I thought about a session I sat in not long ago, going through a current-state assessment with a room full of people who’d each built a version of “how things work” in their own head, none of which quite matched. I had the deck. I had the rationale about why the architecture should be simpler. What I didn’t have was a way of making anyone in that room see what their Tuesday would actually look like six months from now.
Nobody argued with the logic. Nobody needed to. They just didn’t move.
That’s when Johnson’s point landed somewhere it hadn’t before. The gap wasn’t in my analysis. It was that I’d given people a conclusion and skipped the part where they get to live inside it first.
What Changed
I don’t think this means the analysis was wasted, or that I should start every proposal with a parable. That’s not it. What shifted is smaller than that.
I stopped treating the story as the thing you add at the end, for polish, after the “real work” of the architecture is done. The story is where the decision actually happens. The architecture is just the terms it happens on.
Johnson leans on novels to puts you inside someone’s feelings, hesitation, their competing loyalties, the version of the future they’re afraid of. A target-state diagram never does that. It can’t. It wasn’t built to.
So now, before I put a framework in front of a room, I try to find the one specific person whose Tuesday it will actually change, and I tell that story first. Not as a warm-up act. As the argument.
If You’ve Been Doing This Too
If you’ve spent years assuming the right model wins the room on its own merits. I don’t think you’re wrong, exactly. I think you’re just missing a step. The map still matters. But before anyone accepts a map of a future they haven’t seen, someone has to walk them through what it feels like to stand in it.
I’m still working out how to do that well. Some days the story lands and some days it doesn’t, and I don’t have a framework for which is which yet.
But I’ve stopped starting with the slide.