There’s a version of overconfidence that doesn’t look like arrogance. It looks like experience. You’ve done this enough times. You know how it goes. So you stop asking.
I spent most of 2025 building without asking. Not recklessly — I had hypotheses, data from previous projects, strong opinions shaped by years of making similar calls. I had context.
I just didn’t have current signal.
That felt fine. Until it didn’t.
The story I told myself
The reasoning seemed solid.
I’d seen these patterns before. I knew which objections were real and which were noise. I’d been in the room when products failed, when launches flopped, when teams scattered. I had the scar tissue.
So when I made decisions, I made them from that experience. I didn’t poll. I didn’t run calls. I didn’t circulate drafts for comment.
I decided. I moved. I built.
That’s not wrong, exactly. Speed matters. Consensus-seeking can paralyze. Some feedback loops are just delay with extra steps.
But I crossed a line I didn’t notice crossing.
I went from “I don’t need every opinion” to “I don’t need outside input at all.”
What I missed
The problem with building in isolation isn’t that you lose access to good ideas.
You probably have good ideas.
The problem is you lose access to disconfirmation. Nobody tells you when your assumptions are quietly wrong. Nobody flags the thing that stopped mattering six months ago. Nobody says “I tried that — here’s what actually happened.”
You’re working from a map. The map was accurate when you drew it. The terrain keeps moving.
I had one project where I was convinced the bottleneck was visibility. I designed a whole system to fix it. Ran it for three months. Then someone outside the loop asked a simple question: “Have you asked the team what the bottleneck actually is?”
I hadn’t.
Turns out it wasn’t visibility. It was prioritization. Different problem. Different fix.
I’d built the right solution to the wrong question. Three months in.
Why I stopped asking
I can trace it now.
Somewhere in my mid-career I started treating feedback-seeking as a sign of uncertainty. Strong people decided. They didn’t circulate drafts. They didn’t run surveys or hold “feedback sessions.” They backed their read.
That’s a story I absorbed, not a law I tested.
It has some truth in it. Constant approval-seeking is exhausting and often counterproductive. You can’t make every decision by committee. You do need to back your own judgment.
But there’s a version of that belief that becomes a permission slip to stop listening entirely.
I used it that way.
And the cost was slow. That’s what makes it hard to see. It doesn’t show up as one big failure. It shows up as steady inefficiency. Slightly wrong assumptions baked into every decision. Compounding quietly over months.
What changed
I started a practice I now call deliberate small loops.
Not big feedback programs. Not quarterly reviews. Not an open-door policy.
Just one specific question, to one specific person, every week.
- “Does this match what you’re actually seeing?”
- “What would make this useless to you?”
- “What am I optimizing for that no longer matters?”
That’s it. Small surface area. Low friction. Direct.
What surprised me was how often the answer changed something real. Not every time. Maybe one in four. But that one-in-four was worth more than three months of solo iteration.
The feedback I’d been avoiding wasn’t hostile. It wasn’t complex. It just required asking.
The thing I had backwards
I thought experienced people needed less feedback.
The opposite is true.
Experienced people have more embedded assumptions — assumptions that made sense at a different time, in a different context, under different constraints. They need more active recalibration, not less.
The longer you’ve been doing something, the more important it is to regularly ask: is this still the right read?
Not because experience is worthless. Because experience is incomplete. It always is.
What I’m watching now
I still trust my judgment. I still move fast. I still back my own read of a situation.
But I build in small checkpoints now. Not to get permission. To stay calibrated.
The goal isn’t to slow down for input. It’s to make sure I’m moving in the right direction before I travel too far.
One wrong assumption, caught early, costs an afternoon. The same assumption, caught six months in, costs a quarter.
I learned that the expensive way.