Managing up is mostly writing things down
The leaders I have worked for well were not the ones I agreed with. They were the ones who never got surprised by me. What I send, when I send it, and how I disagree without losing the room.
I have worked for a director who wanted a dashboard, one who wanted a paragraph, and one who wanted to be told nothing until it was done. What they had in common was that the moment they were surprised by something on my team, my credibility with them dropped, and it took a quarter to earn back.
That is the whole job of managing up. Nobody above you should ever be surprised by your team. Everything else is style.
The weekly note
The single practice that has done the most for my relationship with every leader I have reported to is a written update, every week, same day, same format. What shipped. What slipped and why. What I need. What I am worried about that has not happened yet.
The last section is the one that matters. Leaders can handle bad news. What they cannot handle is bad news they should have had a month ago. When I write "I am worried the integration will slip because the vendor has not delivered the API spec" in week two, and it slips in week six, I am the person who called it. When I do not write it, I am the person who hid it.
The note is short, under three hundred words, and it goes to my manager and to the team in the same words. If I would phrase something differently for one audience than the other, that is a sign I am managing a narrative instead of reporting.
Bring the decision, not the problem
Leaders are decision machines with a small input buffer. Bringing a problem with no proposed answer uses their capacity for something I should have done. So every escalation I make has the same shape: here is the situation, here are the two or three options, here is the one I recommend and why, here is what I need from you.
Most of the time the answer is "do the one you recommend." That is the point. The leader has been informed, has had the chance to redirect, and has spent ninety seconds on it instead of a meeting.
Disagree in writing, early, and once
I disagree with decisions from above fairly often. The way I do it has changed.
Early, because a disagreement raised after the decision is complaining. I say it when the decision is still open, in the same written form I use for everything else: here is what I think is wrong with this, here is what I would do instead, here is what I am missing if you see it differently.
Once, because a disagreement raised three times is a campaign. If I have made the case and the decision stands, I execute it as if it were mine, and I tell the team it was decided, not that I lost.
The exception is a decision I believe will harm people on my team or the company in a way the leader has not understood. For those I ask for a real conversation, and I say why it is different from the usual pushback.
Learn what they are measured on
Every leader I have worked for was being measured on something, and it was rarely the thing they talked about most. The director who asked about velocity was actually being asked about attrition. The VP who wanted feature dates was being asked by the board about cost.
Once I knew what they were measured on, I could frame my team's work in terms they could use upward. That is not spin. It is translation, and it is most of what makes a manager useful to the person above them.
The rollout, seen from above
When I proposed moving the team to an AI first way of working, I did not bring my leadership a vision. I brought them a two week evaluation with a scorecard, a six week pilot with three named projects and a traditional estimate for each, and a number they would be able to see in the platform analytics. The proposal fit on one page and the ask was small.
When the pilot delivered, the note they got said what landed, what it had been estimated at, and what I recommended next. When adoption stalled at 40% on one team in week five, they heard it from me that week, with the cause and the fix. By the time the rollout reached 80%, they had never once been surprised, and that is why they let me take it beyond engineering.