I once sat in a meeting where we talked about a product issue for longer than it would have taken to build a fix and run an A/B test to see if it even mattered.
We debated intent. We debated edge cases. We debated how users might react.
Nothing shipped.
By the time the meeting ended, we had strong opinions and zero data. That was the moment it became obvious we were optimising for agreement, not progress.
The ownership problem
The room was full of capable people. Everyone had context. Everyone had a view.
No one owned the decision.
When ownership is shared, responsibility becomes abstract. People argue defensively because being wrong feels personal while success feels collective. The safest move is to slow things down.
This is an uncomfortable truth most startups learn the hard way. Ownership only works when it is clear and singular.
That does not mean ignoring feedback. It means one person integrates input, makes the call, and lives with the outcome. Without that clarity, teams drift toward committees. Committees optimise for safety, not momentum.
Decision fatigue and forced opinions
Another thing that kills speed is asking everyone to have an opinion on everything.
In that meeting, people were pushed to take positions on details they would never be accountable for. Once those positions were stated, they had to be defended. The discussion stopped being about learning and started being about consistency.
This is how factions form. Not because people fundamentally disagree, but because they are forced to pick sides too early.
Most decisions do not need universal buy-in. They need a clear owner and enough context to move forward.
Disagree, then commit
Good teams disagree early and briefly. Bad teams disagree indefinitely.
The difference is what disagreement is used for. “This might fail if X happens” is useful. “I don’t agree with this” without ownership rarely is.
Once a decision is made, the job is no longer to keep debating it. The job is to execute and learn. If it fails, that data is more valuable than another round of opinions.
How to fix it
This does not require better meetings or a culture reset. It requires better mechanics.
- Give initiatives a single owner.
- Share context, not votes.
- Escalate real disagreements quickly.
- Make small, reversible bets and test them in production.
If we had coded that solution instead of debating it, we would have had an answer in days instead of opinions in circles.
Alignment over consensus
Consensus feels safe. Alignment creates motion.
Startups rarely fail because they make the wrong decision. They fail because they avoid making decisions at all. Political paralysis does not announce itself. It looks like thoughtful discussion that never turns into action.
The goal is not to be reckless. It is to fail fast on small bets and iterate faster than we can debate.
Execution beats agreement. Every time.