Leadership That Allows Mistakes, Rather Than Preventing Them

Many organizations define stability as the ability to prevent failures.

More tests. More approvals. More control mechanisms.

But in live systems, that’s a mistaken definition.

The difference between a stable organization and a fragile one isn’t in the number of failures - it’s in what happens when they occur.

The Illusion: If We Prevent Mistakes, the System Will Be Stable

The idea is understandable: mistakes are dangerous. A production failure is a failure. So they should be avoided at all costs.

But when you try to build a system where “bad things don’t happen” - you get a system where:

• People hide problems
• Decisions get postponed
• And small changes become dramatic

The system might be quiet - but it’s quiet in a dangerous way.

Why “Zero Failures” Is a Destructive Goal

A zero-failures target creates a clear incentive:

not to reduce risk - but to avoid exposure.

In such organizations:

• Small failures don’t get reported
• Exceptions get “swept under the rug”
• And everyone learns very quickly what not to bring up

The result: the next failure will already be bigger - and less understood.

How Fear Changes Code

Fear doesn’t stay at the human level. It gets etched into the system.

This is what it looks like in code:

• Double protective layers “just in case”
• Tests that prevent change, rather than protecting it
• Complicated code designed not to bother anyone

This isn’t code written to work well. It’s code written not to draw attention.

A Practical Example: A Culture That Tries to Prevent Mistakes

In a certain organization, every small failure led to a harsh investigation.

Not officially - but everyone knew: whoever was connected to the failure would be marked.

Over time:

• Engineers stopped taking initiative
• Changes got postponed “to the next version”
• And failures were handled temporarily, without touching the root cause

The system became stable in appearance - but broken from within.

What Leadership That Allows Mistakes Looks Like

Responsible organizations act differently.

Not because they like failures - but because they understand they’re inevitable.

They:

• Separate responsibility from blame
• Reward early detection, not concealment
• And measure success by the ability to fix - not by the absence of failure

In such organizations, people dare to say: “something here isn’t right” before it blows up.

A Culture That Promotes Fast Fixes, Not Blame

The difference isn’t in the process, it’s in the unwritten message:

• Is it okay to make a mistake?
• Is it okay not to know?
• And is it okay to stop a system in order to understand?

When the answer is “yes” - the system becomes more stable, not less.

The Bottom Line

Stable engineering leadership doesn’t try to build a system where mistakes don’t happen.

It builds an organization where mistakes don’t destroy trust, don’t paralyze change, and don’t accumulate into an explosion.

Looking Ahead

Even when there’s a culture that allows failure, many organizations still don’t truly learn.

In the next post we’ll examine why knowledge exists, but doesn’t move - and how organizations manage to forget the same lessons over and over.