The meeting is going well.
The ideas are sharp.
The debate is honest.
Someone just said something genuinely interesting about user personas.
You’ve now been in the room for two and a half hours.
Outside, your competitor shipped yesterday.
This is the trap.
Not laziness. Not fear.
The trap is that deliberation feels like work.
The whiteboard is full.
The deck is getting better.
Everyone is engaged.
From inside the room, this is indistinguishable from progress.
It isn’t.
At an early-stage startup, the rarest resource isn’t money or talent.
It is learning.
Almost everything that produces real learning happens outside the building:
What your users actually do, not what you thought they’d do
What breaks under real conditions, not controlled ones
What makes someone leave in the first thirty seconds, not the scenario you designed the flow around
Meetings produce alignment.
The market produces information.
Alignment without information is just a shared belief about an uncertain world.
The economist’s term for this is opportunity cost.
Every hour debating a decision carries the invisible cost of everything you didn’t learn because you weren’t in the market.
At a lean startup, that cost compounds fast. You have so few hours before the runway runs out.
Jeff Bezos put the math plainly:
Make decisions with 70% of the information you wish you had. Waiting for 90% is too slow.
At a two-person startup trying to find product-market fit — the point where what you’ve built matches what the market actually wants — waiting for 90% isn’t slow.
It’s a different game.
The nuance worth keeping: not all decisions carry the same cost of being wrong.
Amazon calls it the two-door rule:
Some decisions are reversible. You walk back through the door, adjust, and go again. → Make these fast
Others are one-way. → These deserve more deliberation
Bias to action lives in the first category.
And in early-stage product work, most decisions are in that category.
In practice:
You build a hypothesis.
You put something in front of a user.
You let the feedback tell you what to debate next.
You don’t debate before building.
You debate after.
The market is a better sparring partner than any meeting room.
I see this now with AI development tools.
Claude Code. Codex. Gemini. Cursor.
Teams that should be the fastest in history are instead running workshops on which tool is philosophically superior.
The prompt engineering debate replaced the architecture debate,
which replaced the framework debate before it.
The object of the argument changes.
The argument never does.
While that conversation is happening, someone else:
Picked a tool
Shipped ten features
Learned which three worked
Moved on to the next problem
They don’t have a better process.
They have fewer opinions about the process.
That’s the advantage.
The founder’s structural advantage in a small team has always been speed:
Fewer layers
Faster decisions
Direct contact with the market
The teams that waste it aren’t lazy.
They’ve confused the comfort of a well-run meeting
for the signal of a real decision.
Tobi Lütke framed the job this way:
You don’t earn your place by knowing what to do.
You earn it by making good decisions when you don’t know what to do.
The action is the research.
Ship it.
Then find out what you actually built.