How to Manage Up
Calibrating the One Relationship You Cannot Get Wrong

The key insight:

The question is never whether to disagree with someone who can end your tenure. It is how many times you have earned the right to, and whether this is one of them.

Most content on managing up is written for someone managing a line manager: a peer relationship with a title difference, where the worst outcome is an awkward review. That is a real skill, and it is not the one this article is about. This is written for the version that shows up once your boss is a CEO, a founder, or a board you report into directly, where the relationship carries a structural asymmetry no amount of goodwill removes: they can end your tenure, and you cannot end theirs.

That asymmetry changes the calculation on almost everything: how much certainty to project, when a problem is yours to solve quietly and when it needs to be surfaced immediately, and how to disagree with someone whose good opinion of you is not a nice-to-have but the actual condition of your continued employment.

The skill is calibration, not deference

The common failure at this level is not usually open conflict. It is drifting toward one of two extremes: either performing certainty you do not have, because admitting doubt to someone who can end your tenure feels dangerous, or performing agreement you do not hold, because disagreement feels riskier than it actually is. Both responses are attempts to manage the relationship's risk, and both quietly erode the thing that makes you useful to a powerful boss in the first place: an honest, calibrated read they can actually rely on.

The useful frame is not "how do I keep this person happy." It is "how do I remain the source this person trusts most when the information is bad," because that is the actual value a senior report provides to someone at the top of an organisation, and it is precisely the value that performing certainty or performing agreement destroys.

Naming your confidence level explicitly

Most senior professionals learn, correctly, that vague hedging reads badly to a CEO or a board. The overcorrection is projecting more certainty than the evidence supports, which works until the first time reality disagrees with the confident version, at which point the cost lands all at once and is disproportionate to the original uncertainty.

The fix is not less confidence. It is naming your actual confidence level as information, rather than letting your tone imply a level you have not earned. "I am confident in this number, we have three quarters of consistent data" and "I believe this is directionally right, but we are working from one data point" are both more useful to a boss who has to make decisions on what you tell them than a uniformly assured delivery that turns out to have been guessing on the parts that mattered.

Disagreeing with someone who can end your tenure

The honest question is never whether to disagree. It is how much credibility you have banked for this specific disagreement, and whether the issue is significant enough to spend it on. A senior professional who disagrees with everything reads as difficult regardless of how right they are. One who never disagrees becomes someone whose agreement carries no information, because it was never actually being withheld as a genuine option.

The useful discipline is picking the hills that matter and being visibly, consistently right about the ones you choose to stand on, which is what actually buys the credibility to be heard the next time. The Center for Creative Leadership's ongoing research programme studies exactly this kind of judgement under pressure and uncertainty as a core, learnable leadership capability, rather than treating it as an innate trait some people simply have and others do not, which is the more useful frame than generic assertiveness advice.

Why this is not the same skill as stakeholder management or influencing without authority

Two adjacent pieces in this collection cover related but genuinely different ground, and the distinction matters because the fixes do not transfer. Stakeholder management is the standing system across many relationships, most of which carry no formal power over your role at all. Influencing without authority is peer-level persuasion, moving an equal toward an outcome, with no power asymmetry running either direction. Managing up is neither. It is one specific relationship, upward, with someone who holds real power over your tenure, and the asymmetry itself, not the number of people involved and not the absence of formal authority, is what makes it a distinct skill.

CIPD's factsheet on management development treats management competency as something to be identified and developed deliberately, rather than assumed to be present once someone is promoted, which is the right instinct to apply to the upward relationship specifically. Most development plans invest heavily in managing downward and almost nothing in managing upward, despite the upward relationship being the one with the most direct bearing on whether a career actually advances.

This same asymmetric-escalation judgement has a named governance analogue. JOH Partners' research on the CHRO's board-anchored line of sight looks at how a senior function head decides what genuinely needs to reach the board directly versus what should be resolved inside the function first, the identical calibration this article describes for an individual relationship, formalised as a standing reporting line.

If you want a structured read on where your own capacity for this kind of calibrated judgement sits, the free Leadership Psychometric on AssessYou scores Influencing Power alongside Vision Thinking, Execution Drive, People Intelligence and Composure. Take the free Leadership Psychometric to see your profile, or read the Leadership Psychometric overview first for a look at what the report covers.

Key takeaways

• Managing up at senior level means calibrating communication with someone who has real power over your tenure, a structural asymmetry that does not exist with a peer.
• The senior failure mode is rarely open conflict. It is drifting toward performed certainty or performed agreement, both of which erode the honest, calibrated read that makes you valuable in the first place.
• Name your actual confidence level explicitly rather than letting tone imply a level you have not earned. Uniform confidence is not the same skill as accurate confidence.
• Decide case by case whether an issue is yours to own or theirs to know about. Getting that judgement wrong in either direction usually costs more than the underlying issue.
• Managing up, stakeholder management and influencing without authority are three distinct skills. They are frequently confused because all three get grouped under "political skill," but the fixes do not transfer between them.

For more on the standing relationships and capabilities senior careers actually turn on, see the full Career Strategy collection, including our guide to mapping the realistic routes above you, the next step once the relationship itself is well managed.

JOH Partners Intelligence

Get the fortnightly brief on GCC careers, salaries, and market moves.

No spam. Unsubscribe any time.