You can run a flawless team and still watch your work stall, because the decision sat with a boss you hadn't briefed, or a peer in another function who had no reason to prioritise you. The instinct is to call this "politics" and resent it. The more useful read is that influence travels in three directions, and the two you don't control, up to your manager and across to your peers, are where most careers quietly succeed or fail.

The quick version

  • Managing up is the deliberate work of building a productive relationship with your boss, not flattery, but understanding their goals, pressures and working style so you can do your job well and make theirs easier.
  • Managing sideways (or "across") is getting cooperation from peers you have no authority over, the colleagues in finance, legal or another team whose help you need but cannot demand.
  • Both run on the same engine: reciprocity. You earn cooperation by understanding what the other person needs and trading something they value, not by relying on rank you don't have.
  • The trap is hearing "managing up" as manipulation. Done honestly it's the opposite, it's making the relationship work in the open, so the right things get the resources and attention they deserve.

The idea in depth: your boss is a relationship, not a fixed obstacle

The foundational text here is older than most readers and still the sharpest. In "Managing Your Boss", first published in Harvard Business Review in 1980 and reissued as a Classic in January 2005, John Gabarro and John Kotter make a claim that still lands as mildly subversive: you are responsible for managing the relationship with your boss, not just receiving instructions from it. Their core argument is mutual dependence. Your boss depends on you for honesty, reliability and information from the front line; you depend on them for priorities, resources and air cover. Neither side can do their job without the other, yet most people manage only the downward half of that relationship and leave the upward half to chance.

Gabarro and Kotter's practical instruction is to understand your boss as a real person under real pressure: their goals and pressures, their strengths and blind spots, and, unglamorous but decisive, their preferred way of taking in information. Some bosses are readers who want a memo before the meeting; some are listeners who want to talk it through. Get that wrong and a good proposal dies on delivery.

So the move is: in your next one-to-one, stop reporting status and start managing the relationship. Ask two questions you probably never have, "What's the most important thing on your plate this quarter?" and "How do you prefer I bring you things, written ahead, or talked through?" Then route your work to match. You are not sucking up; you are removing the friction that was quietly costing you decisions.

You manage your boss for an unglamorous reason: to do your own job well. The relationship is part of the work, not a distraction from it.

An honest limitation. "Managing your boss" assumes a boss worth managing, reasonable, if flawed. The model is far weaker against a genuinely abusive or chaotic manager, where the same techniques can read as enabling. Gabarro and Kotter were describing how to make an ordinary, imperfect relationship productive, not how to survive a toxic one. If the issue is mistreatment rather than mismatch, this is the wrong tool, that's a question for HR or your own exit, and worth checking against your organisation's policy and your jurisdiction.

The idea in depth: sideways runs on currencies, not commands

Managing across is harder than managing up, because at least your boss is paid to engage with you. A peer in another department owes you nothing. The clearest model for this is Allan Cohen and David Bradford's influence-without-authority framework (developed in their book of the same name and summarised in the Journal of Organizational Excellence, 2005). Their starting point is the Law of Reciprocity: across cultures, people expect to be paid back for what they do for others. Influence, then, isn't something you have, it's something you build by trading in currencies, their word for whatever the other person actually values.

Those currencies are rarely money. Cohen and Bradford group them into families: inspiration-related (a sense of meaning, a worthy cause), task-related (resources, information, a quick "yes" that unblocks them), position-related (recognition, visibility, a good word to their boss), relationship-related (trust, a listening ear), and personal (gratitude, autonomy, feeling respected). The skill is diagnosis: figuring out which currency a given colleague is short of, and whether you can supply it.

flowchart TD
  A(["You need a peer's
cooperation"]) --> B(["See their world:
what pressures them?"])
  B --> C(["Find the currency
they value"])
  C --> D(["Offer it first
(give before you ask)"])
  D --> E(["Make the request
clearly"])
  E --> F(["Cooperation earned
relationship banked"])
  C -.->|"misread the currency"| G(["No traction,
diagnose again"])
  G --> B
Cohen & Bradford's exchange model: influence is earned by trading in whatever a peer genuinely values. Leaders Loop

So the move is: before you ask a peer for anything, spend five minutes answering one question, what is this person measured on, and what would make their week easier? Then lead with that. Send the data they've been chasing, credit them publicly to their manager, or simply ask what's on their plate before you pile on yours. Give before you take, and the request that follows stops feeling like an imposition.

This is more than a coordination skill; it seems to track with career outcomes. In "Why effective leaders must manage up, down, and sideways" (McKinsey Quarterly, April 2017), Thomas Barta and Patrick Barwise argue from their research that strong team leadership alone, managing down, is not enough; the leaders who have business impact also mobilise their bosses and their peers. Their honest caveat is worth keeping: this is correlational and self-reported, the usual limits of survey-based leadership research, so read it as a strong steer rather than a proven law. The direction, though, is hard to argue with, the lateral and upward relationships are not a soft extra, they're where a lot of the work actually gets done.

A worked example

Take a product manager, call her Priya, who needs a security review signed off before a launch. (Illustrative scenario; the figures and names are a teaching example, not a real case.) The review sits with a security lead, Sam, in another department who doesn't report to her, is buried under audit work, and has no deadline that says "Priya's launch." Her first three emails go unanswered. The temptation is to escalate, to ask her boss to lean on Sam's boss.

Run it through the two models instead. Managing up first: rather than asking her director to apply pressure, Priya briefs them the way Gabarro and Kotter advise, a three-line written summary (her director is a reader) naming the one risk that matters: a launch slip. She's not offloading the problem; she's giving her boss the context to help if needed, in the form they digest fastest.

Then managing sideways: she stops treating Sam as a gatekeeper and asks what currency he's short of. It turns out he's drowning and what he values most is a task-related one, a request that's easy to action. So instead of "please review my launch," she sends a pre-filled checklist, the exact files, and a specific 30-minute window. She's spent her effort to save his. The review clears in two days.

flowchart LR
  A(["Blocked: security
review stuck"]) --> B{"Escalate by force,
or earn it?"}
  B -->|"Force: boss leans
on Sam's boss"| C(["Review done late,
Sam resents you"])
  B -->|"Up: brief director
in their style"| D(["Air cover ready,
not yet spent"])
  B -->|"Sideways: supply
Sam's currency"| E(["Easy 'yes',
review in 2 days"])
  D --> E
  E --> F(["Launched, and the
relationship is stronger"])
Same blocker, two routes: force burns the relationship; trading in currencies banks it. Leaders Loop

The escalation route might also have worked, once. But it spends a favour and leaves Sam dreading your name. The reciprocity route costs Priya more effort up front and earns her a colleague who'll move fast for her next time. That asymmetry is the whole argument: pressure is a one-shot currency; reciprocity compounds.

Frequently asked questions

Isn't "managing up" just a polite word for sucking up?

It's the opposite when done honestly. Flattery serves you and hides information; managing up serves the work and surfaces it, including bad news your boss needs early. Gabarro and Kotter frame it as making a relationship of mutual dependence function, so the right priorities get the right resources. If your "managing up" involves telling your boss only what they want to hear, you've stopped managing up and started managing your own comfort.

How is managing sideways different from just being collaborative?

Collaboration is the goal; managing sideways is the deliberate method when goodwill alone isn't enough. With a peer whose priorities genuinely compete with yours, "let's collaborate" is too vague to move them. The Cohen–Bradford lens makes it concrete: identify what that specific person values, offer it, then ask. It's collaboration with a diagnosis attached.

What if my boss is the problem, disengaged, or worse?

Match the tool to the situation. For an ordinary, busy, imperfect boss, the managing-up moves work: understand their pressures, adapt to their style, bring solutions not just problems. For a genuinely abusive or unsafe manager, these techniques won't fix it and can quietly make you complicit, that's a matter for HR, documentation, and possibly your exit, guided by your organisation's policy and local employment law rather than a leadership framework.

I have no time for relationship-building. What's the minimum that works?

Two habits, both cheap. With your boss: once, ask how they prefer to receive information, then route everything that way, it saves you both time on every interaction afterward. With peers: before any ask, spend two minutes naming what they're measured on and lead with something useful to it. You're not adding relationship work to your week; you're making the work you already do land the first time instead of the third.

Does this still matter if I'm senior or a CEO?

More, not less. Barta and Barwise found senior leaders still need to mobilise upward (to a board or chair) and sideways (to C-suite peers whose cooperation they can't command). The higher you go, the more of your job is done through people you don't control, which is exactly why managing up and sideways scales with seniority rather than fading out.

Related in the Toolkit

Managing up and sideways rests on two things you have to do inside yourself first: knowing how you come across (self-awareness & reflective practice) and choosing where your limited effort goes (prioritisation & focus), since you can't trade currencies in every direction at once.

Where to go next