Disagreement Is Data

A few years into Verizon's business onboarding redesign, I hit a wall that had nothing to do with information architecture. It was about paint colors.

Specifically: where should a device's color options live? My position was that colors belong with the device's specs — screen size, storage, camera — because that's how people actually decide on a phone. Color is part of the object, not an accessory to it. The team's position was different: colors sat in a separate step, grouped with add-on services like cases and screen protection.

I was sure I was right. I was also sure that being sure wasn't the same as being correct.

The instinct in that moment is to defend your position. The useful instinct is to ask why theirs exists at all.

So before I made my case, I asked why the other structure existed. It turned out there was real reasoning behind it — the add-on flow had its own conversion goals, and grouping color there was thought to increase engagement with the whole step. That's not a naive idea. It's a defensible one.

I still disagreed. But now I disagreed with a specific claim instead of a vague feeling, which meant I could actually test it. I laid out how people build a mental model of a device — color first, because it's the first thing they see and the thing they'll live with — and argued the add-on step should earn attention on its own terms, not borrow it from color choice.

The decision went the other way. Not because my reasoning was dismissed — the team weighed both, and theirs won that round. I didn't get the outcome I wanted. What I did get was more useful than winning: a real test of whether my model of the user was right.

Months later, working through a related update, the client moved the structure toward what I'd originally proposed. Nobody made a moment of it. There was no "you were right" conversation. The interface just quietly changed shape, and I recognized it when I saw it.

If you're right, the evidence tends to show up eventually, with or without your name on it. If you're wrong, you find out fast — which is its own kind of win.

Here's what I actually took from it. Disagreement isn't a threat to a decision — it's a free stress test of one. Every time someone pushes back with a real reason, you get to check your thinking against theirs before either of you has to live with the consequences. Treat it as a fight to win in the room, and you're optimizing for looking right today. Treat it as data, and you're optimizing for actually being right, on whatever timeline that takes.

The onboarding flow this argument was part of eventually went from 9 steps → 4. Nobody remembers the color debate specifically. But it's the reason I trust the rest of the decisions in that flow — because they all survived the same test.