Engineering leadership

Leadership problems are often paradoxes

Good leadership is often not choosing one side of a tension. It is making both sides explicit enough that people can operate.

2026-06-08 Adapted from a LinkedIn post
Back to writing

The hardest leadership problems are often not choices.

They are paradoxes.

Parenting taught me this before management did.

We want children to listen to us. We also want them to become people who can think without us.

We want them to inherit some of our values. We also want them to accept people who are not like them.

And maybe the great paradox of parenting is this: you are trying to raise someone who can eventually leave you.

Leadership has the same shape.

We want engineers to pick up the highest-priority work, even when it is not theirs. We also want smaller teams to feel real belonging and ownership.

If every boundary is sacred, priorities fragment. If every boundary is optional, ownership becomes theater.

We want autonomy. But autonomy does not mean silence, isolation, or refusing help.

For a new engineer, autonomy may look like: I tried this, observed this result, and I need help choosing between these two next steps.

That is not dependency. That is structured judgment.

We want candor. But candor without care becomes ego.

We want kindness. But kindness without truth becomes avoidance.

We want leaders to make decisions. Then, when the decision affects our work, we may call it micromanagement.

Sometimes we want consensus because we value inclusion. Sometimes we use consensus because nobody wants to own the disagreement.

The work is not to resolve all of this into one clean rule.

The work is to name both sides clearly enough that people can operate.

Ownership needs boundaries. Autonomy needs escalation paths. Candor needs tact. Kindness needs standards. Delivery needs people. People need delivery to matter.

I distrust leadership advice that makes one side sound virtuous and the other side sound backward.

Real teams usually need the AND.

The failure mode is not paradox.

The failure mode is pretending there is no paradox, and then blaming people when they optimize for the wrong half.

Which leadership paradox became easier for you only after both sides were named explicitly?