When something goes wrong in an organization, the instinct is to find the decision and the person who made it. The postmortem locates the moment of choice. The question becomes: why did they decide that?
This framing is almost always wrong. Not because people are blameless, but because individual decisions are the last observable step in a chain of prior constraints. By the time someone chooses, the outcome has mostly already been shaped.
The information problem
Every decision runs on information. Most information reaches the decision-maker through channels that were designed for something else — a report format inherited from a predecessor, a dashboard built for a different question, a meeting structure that predates the current problem.
When the information is late, incomplete, or shaped by the wrong categories, the decision is not a failure of judgment. It is a failure of infrastructure. The person at the end of the chain is being asked to compensate for a system that never delivered what they needed.
The default problem
Most decisions are not active choices. They are defaults accepted without friction. The path of least resistance becomes the path taken — not because it was selected, but because alternatives required more energy, more information, or more permission than was available in the moment.
An organization that wants better decisions should spend less time training judgment and more time making the correct path the default path. When the right thing is also the easy thing, decision quality stops depending on heroism.
The feedback problem
Decisions improve when outcomes return to the decision-maker. But most organizational feedback loops are broken. The person who approved the budget does not sit through the project that burned through it. The person who chose the vendor does not operate the tool daily.
Without feedback, decisions drift. Patterns repeat. The same class of mistake recurs under different names. The fix is not more training. The fix is closing the loop: making sure whoever decides also sees what happened.
What to build instead
A decision-making system — as opposed to a collection of individual decisions — needs three things:
- Clean information flow. The right data, at the right time, in a format that answers the actual question being asked.
- Deliberate defaults. The easiest path should be the one you want people to take. Design defaults, do not just hope for good choices.
- Closed feedback. Every significant decision should return its outcome to the person who made it. Not as blame, but as calibration.
None of this requires smarter people. It requires better systems. The kind that make good decisions easy, bad decisions hard, and every decision visible enough to learn from.
The person who appears to have excellent judgment usually has excellent systems feeding them. The person who appears to have poor judgment is often just downstream of broken ones. Fix the system, and the decisions fix themselves.
Back to writing