[organsnyder]

> strongly consider keeping your mouth shut if that way is only 5% better

That's great advice. An experience with a director overruling his engineering staff is one reason I no longer work for him.

As a leader, your most valuable asset is context: big-picture strategy, business priorities, and the like. If you're in an interaction where you find you're not utilizing that asset, you probably should back off and let your reports do their jobs (unless you're mediating a dispute, but even there you should help people work together better—not just pick a side).

[segmondy]

If the director is smarter and more knowledgeable about the engineering domain, he should overrule. that's what the big-picture strategy is all about, some people make it into leadership because they are really good and capable at engineering. So long as they stay sharp, they can still make technical decisions.

[ska]

Not really, in my experience. If it’s a real director position rather than in name only you simply don’t have time to get into the details enough to be good at this - if you have enough context you aren’t doing your real job well. This can feel like positive contribution but you are actually slowing down the entire group.

What you should be doing is focusing on making sure the decision making process is working well and that you are all learning from issues. If you have an idea you think is better, by all means hand that off to the team. But you have to let them decide against it (so long as they can articulate why). Absent strong empirical evidence that the path is wrong, you go with their choice and discuss how to validate it.

In all cases except team incompetence you are better off this way in the long run. And if the team is incompetent, you have bigger problems.

[organsnyder]

That might lead to the best technical outcome (if the director is actually right!). But the message it sends to subordinates is incredibly damaging. If a leader really needs to pull rank because they can't explain why a different approach is worthwhile, either they're not a good communicator, and/or their subordinates are incompetent. And this discussion should only be occurring for major, irreversible decisions in the first place.

[weej]

Agreed. A leader needs to be capable of indirectly & positively influencing the team's architecture and overall technical outcome while not doing so explicitly (manage down, up and out). The team needs to drive and come to their own conclusions, but a true leader can positively impact the vision and strategic execution.

[wsgreen]

Everyone thinks they're smarter or more knowledgeable and will then always jump in