Managing Up Is Not Flattery — It Is Designing the Flow of Information

Introduction
1. The Resolution at Which Your Manager Can See You
2. Three Things That Cause Trouble If Your Manager Does Not Know
3. You Set the Reporting Cadence
4. What Your Manager's Manager Is Graded On
5. Building a User Manual for Your Manager
6. Working With a Bad Manager
7. When to Stop Trying
8. One Thing You Can Do Today
Where This Advice Does Not Apply
Closing
Further Reading

Introduction

Managing up has a bad reputation. The words that come to mind first are flattery, reading the room, picking a side. So people who intend to win on merit skip the whole area.

The results of skipping it are generally not good. You worked hard on the wrong thing. You did well and nobody heard about it in time. You spent two months going in the opposite direction from something that had already been decided upstairs. These repeat. This is not a problem caused by failing to flatter anyone. It is a problem caused by information flowing in only one direction.

This is not a post about getting along with your boss. It is about designing what information moves between the two of you, in which direction, and how often. A better relationship is a by-product, and even when the by-product does not appear, the value of the design remains.

1. The Resolution at Which Your Manager Can See You

Start with the structure. One manager watches several people. Following every person's work in full is physically impossible, so a manager collects partial signals and reconstructs the whole from them.

Where do those signals come from? The traces you leave, what gets said in meetings, remarks other people make in passing, and the noise when something blows up. What stands out in that list is that work that is functioning normally produces no signal.

So a structural distortion appears. Work that is running well is invisible to your manager, and only what broke is visible. And good news arrives inflated through several paths while bad news arrives late and shrunk, because each person passing it along shaves a little off.

The reverse direction is the same. I see my own work deeply, but I do not know what constraint got applied upstairs or how priorities shifted at yesterday's executive meeting. So what I miss is also the same kind of thing every time.

In one sentence: your manager sees wide but shallow, and you see narrow but deep. So what your manager misses is always the detail coming up from below, and what you miss is always the constraint coming down from above. Filling those two holes is the whole of what this post calls design.

2. Three Things That Cause Trouble If Your Manager Does Not Know

You cannot report everything. That eats your manager's time, and eventually nothing gets read at all. So you need a criterion for what to let flow.

There is one criterion. Would it be a problem if your manager learned this from a path other than you first? If yes, send it. If no, you do not have to. Three kinds actually catch on this.

First, whatever will get asked from above. Your manager also fields questions from their manager. With no answer, your manager is put on the spot right there, and the memory of being put on the spot attaches to you. So learn which of your items the level above cares about, and send the status of those items before they are asked for.

Second, whatever affects the schedule. The problem is less the slip itself than learning about the slip late. Your manager holds several schedules interlocked, so when one moves, several things need adjusting. With time to adjust, a delay is not an incident. Without it, it is.

Third, friction between people. A conflict with a neighboring team, a request that got turned down, someone who is not cooperating — these are generally the kind of problem a manager can unblock. But reporting it feels like telling on someone, so people put it off, and when they do, the other side goes up first. The narrative that goes up first becomes the default.

Things outside those three are generally fine inside the periodic report.

3. You Set the Reporting Cadence

Most people report when they are asked. And the ask comes after the other person has started to wonder. So a report driven by a request is structurally always late.

Declaring the cadence yourself changes three things.

First, the meaning of silence changes. If you have set Tuesdays, then silence until Tuesday is not anxiety but schedule. This matters more than it looks. A large share of the cognitive resource your manager spends on you is burned in the state of not knowing how it is going right now.

Second, check-in requests decrease. A sudden message asking about progress is usually an expression of unease, and when a report is already scheduled, that unease never forms.

Third, it becomes easier to keep. With a set point you just write at that point; without one, you have to judge each time whether to send now. That judgment tilts a little toward later every time, and eventually you stop sending.

The shorter the format, the better. What is done, what is in progress, what is blocked, and what needs your manager to decide. Four lines is enough, and past four lines it does not get read. The last item matters most. If you do not state that a decision is needed, the report gets read and nothing happens.

And if you have a one-on-one, do not fill that time with progress. Progress already went out in the periodic report. The other uses of a one-on-one are covered in How Not to Waste a One-on-One.

4. What Your Manager's Manager Is Graded On

This is where it starts to sound like flattery, so let me draw the line first.

Knowing what your manager's manager is graded on is not about looking good to that person. It is about knowing which language your work has to be described in for it to travel upward. When the same outcome can be described two ways, choosing the one that overlaps with the language used above is translation, not distortion. It is a different thing from manufacturing an achievement that does not exist.

Look at the structure and the reason is clear. Your manager's judgment about your work is not made by your manager alone; it is influenced by whether the work takes a shape your manager can explain to their own manager. An achievement that is hard to explain does not travel upward even when it is a good one. That is a loss for your manager and for you.

There are three ways to find out, and they have an order.

First, public documents. Org goals, quarterly plans, all-hands material. The words that recur there are what the organization is currently counting. Remarkably, a lot of people never read them.

Second, the words your manager repeats. People unconsciously repeat the axis they are graded on. If there is a concept your manager brings up three or more times in meetings, it is likely the axis that came down from above.

Third, the questions that make your manager tense. Watch which topics make the answers get longer and more defensive, and you can see where that person is currently under pressure.

The reason observation comes first is the same as what is explained in a later section. In Korean organizations, checking directly by going around the line is far more expensive than observing.

5. Building a User Manual for Your Manager

People work differently, and the difference is habit rather than personality. Habits can be learned by observation, and knowing them cuts friction sharply. Roughly these are the items worth writing down.

Channel. Where urgent things go and where non-urgent things go. Some people do not read chat and some do not read email.
Response time. Within how many hours an answer usually comes. Knowing this lets you decide whether to resend or wait when there is no reply.
Document shape. Whether this is someone who reads long documents, or only looks at a three-line summary, or likes tables, or prefers to hear it spoken.
Delegation boundary. What you are told to handle on your own and what you must run past them. Misread this line and you become either high-handed or exhausting.
Decision rhythm. Whether this is someone who decides on the spot or sleeps on it. Demand an answer on the spot from the second kind and it is usually no.

These are all things you are allowed to ask about. Just after joining, or just after a change of manager, is an especially good moment to ask, and the question itself reads as a signal that you are someone who designs their work.

6. Working With a Bad Manager

The common mistake here is treating a bad manager as one category. Different kinds need different responses. In practice four of them show up mixed together.

Incompetence. Judgments are frequently wrong and direction changes frequently. The response is records. Put decisions and their timestamps in a document, and when direction changes, record the change too. Not to fight, but so that when direction changes a third time, the changing itself can be raised as a subject.

Neglect. No feedback and no interest. This looks comfortable and is the most dangerous. The path by which your work travels upward is blocked outright. The response is section 3: you set a cadence and push it in. If there is still no reaction, you need to build another path inside the organization.

Credit-taking. Your results are conveyed upward as your manager's. The response is not an expression of feeling but traces. Put author and date on documents, share results in the team channel, and shape them so that people on other teams can cite them. This direction is covered in more detail in Being Visible Inside the Organization.

Public reprimand. Being cut down in front of people. The first step of the response is to move the setting. Pushing back on the spot is usually a loss. Later, ask for a separate moment, agree with the substance of the criticism, and say one explicit thing about the delivery. This one time has to happen. Without it you cannot apply the criteria in the next section.

7. When to Stop Trying

At some point you have to decide whether to keep trying to improve things. Trying indefinitely is not conscientiousness but loss. I would suggest three criteria.

First, have you asked for the change explicitly even once? In words, not through hints or attitude. Concluding that someone is a bad manager without having done this is early. People genuinely often do not know how their behavior is landing.

Second, what changed in the quarter after that? You are not expecting a complete transformation. You are looking at whether even one item moved. If nothing moved, nothing will move going forward.

Third, does your reputation flow only through this one person? This is the most important criterion. If the only person in the organization who knows you is your manager, then that one relationship is your career. If people on other teams know your work, the situation is far better.

If the answer to the third question is not good, the thing to do before leaving is to add paths. And if that work makes no progress over several quarters, you move to a different calculation then. When to Leave and When to Stay is set out separately.

8. One Thing You Can Do Today

Add one question to your next one-on-one or your next conversation.

Out of the things I work on, which one do you get asked about most often from above?

That single question stands in for several of the sections above. The answer tells you at once what to send before it is asked for, which language to explain it in, and what is currently being counted in the organization.

And if your manager cannot answer it, that is information too. It means either your work is not an object of interest above, or your manager is not connected upward. Both are worth knowing.

Where This Advice Does Not Apply

Harassment, discrimination, and retaliation are not management problems. They are safety problems, and trying to apply the methods in this post there costs you time. What is needed is not information-flow design but records and paths outside the reporting line. Write down what happened and when, factually, and start by finding the internal reporting procedure, the HR or labor relations function, and if needed an external counseling channel. Advice to try to improve things must not be carried over into this territory.

In Korean organizations the cost of going around the line is different. Advice to ask your manager's manager directly what they care about assumes a flat hierarchy. In an organization with a clear gyeoljae line — the formal approval chain a decision travels up — contact that skips the line is itself an incident, and the cost of that incident exceeds the information gained. That is why section 4 sets observation and public documents as the default. Asking directly is for when your manager creates that setting, or when there is evidence that the organization genuinely encourages such contact.

Under dual reporting and in a matrix organization, the design in this post does not simply double. It gets far more complicated. When two managers have conflicting priorities, absorbing the conflict and reconciling it alone is the most common and the worst response. There is one principle. When you find a conflict, raise it in a setting where both people are present. If you translate in the middle, only the responsibility for the outcome stays with you.

When it is certain your manager is about to change, there is no payback period on the investment. If a reorg is scheduled or your manager is about to move, the time spent designing the relationship is better spent elsewhere. That said, the periodic report and the records carry over even when the manager changes, so those two are worth continuing.

One honest limit at the end. The methods in this post work well when your manager is not a bad person. A problem caused by missing information is solved by information, but when the problem is intent, giving more information does not improve it. Sometimes it supplies ammunition. So the first judgment in this post is always to distinguish whether the problem you are having is a problem of information or a problem of intent.

Closing

The reason managing up gets mistaken for flattery is that people assume the object of the verb is a person. The object is not a person but information. It is deciding for yourself what moves when and in which direction.

And the benefit of this design is not a better relationship; it lies elsewhere. Knowing what is being watched above keeps you from spending two months on the wrong thing. Having your state always visible means requests get filtered before they reach you. Sending bad news up early leaves time to respond.

This design carries over when your manager changes, because it is an investment in a method rather than in a person.