This lesson is the vocabulary the rest of the course spends. Grove wrote it about middle managers at Intel in 1983, and it is the cleanest statement of why an executive’s calendar is the job rather than an obstacle to it.
Grove asked a group of middle managers what a manager’s output is. They said: judgments and opinions, direction, allocation of resources, mistakes detected, people trained, courses taught, products planned, commitments negotiated.
He rejects all of it. Those are activities. They describe what managers do while trying to create a result. The result is something else.
If you run a wafer fabrication plant, your output is completed, high-quality wafers. If you are a high school principal, it is educated students. If you are a surgeon, it is a recovered patient. In each case the manager’s own hands-on work is real and important and is not the output.
Grove’s analogy: a coach or a quarterback does not score touchdowns. League standings are kept by team, not by individual.
The second half of the definition matters as much as the first. Neighbouring organizations under your influence count. A lawyer who gets a regulatory permit releases years of research to the public. An analyst whose interpretation guides the whole company has produced output far beyond her own team. Grove goes as far as saying individual contributors who gather and spread know-how should be counted as middle managers, because they exert great power in the organization.
If output is the organization’s, the question becomes what a manager does to increase it.
Grove writes it as an equation. Managerial output equals L1 x A1 plus L2 x A2 and so on, where each A is an activity and each L is its leverage. Every activity you perform increases the organization’s output by some amount, and the leverage is how much.
That gives three, and only three, ways to raise your productivity.
Working faster is the option everyone reaches for, and it is the one bounded by the number of hours you have. Raising an activity’s leverage, and shifting the mix toward higher-leverage activities, have no such ceiling. Which is why Grove says the key to high output is being sensitive to the leverage of what you do during the day.
Grove gives three sources.
Many people are affected by one manager. His example is Robin, an Intel finance manager who sets up the annual planning process. By defining in advance exactly what information is gathered at each stage and who is responsible for what, she directly affects the work of about two hundred people.
A brief, well-focused set of words or actions affects someone’s behaviour over a long period. A performance review is a few hours of work that can shape a person’s work for a year. So is a five-minute correction given at the right moment.
A large group’s work is affected by one unique, key piece of knowledge. The person who knows the one thing that unblocks everyone else, and says it.
Timing is not separate from leverage, it is part of it. Robin’s work done before the planning meeting has enormous leverage. The same work done later, scrambling to help one manager define milestones, has much less. Grove’s other example is learning that a valued person has decided to quit: act immediately or the chance is gone.
The equation runs both ways, and this half is what makes it useful to an executive.
| Activity | What it costs |
|---|---|
| Arriving at a meeting unprepared | The meeting’s time, plus the time those people could have spent elsewhere |
| A visibly depressed manager | Spreads through the organization without them noticing |
| Waffling on a decision | Work stops, because no green light is a red light |
| Managerial meddling | The subordinate takes a much more restricted view of what is expected of them |
Arriving unprepared at a meeting costs twice: the time wasted and the time those people could have spent elsewhere.
A depressed manager infects the organization. Grove’s example is a manager who saw his division would not make money despite a year of cost reduction, became depressed, and spread it through his organization without realising, until someone on his staff told him what he was doing.
Waffling is putting off a decision that affects other people’s work. The lack of a decision is the same as a negative decision. No green light is a red light, and work can stop for a whole organization.
Meddling is the one worth sitting with, because it is the trap that catches technically strong executives specifically. It comes from exploiting superior work knowledge, real or imagined. A senior manager sees a bad trend and hands the responsible person a detailed set of actions.
The cost is not the one intervention. After repeated exposure, the subordinate takes a much more restricted view of what is expected of them. You have not solved a problem, you have shrunk a job.
Grove notes that a bad training session can be fixed by retraining the group, but depression and waffling are very hard to counter, because their impact is both pervasive and elusive. There is nothing specific to redo.
The CTO who rewrote the deployment pipeline performed a high-quality activity with a leverage of roughly one. She produced her own two weekends of work, once.
The unanswered question, how the company decides what to build next, is Robin’s planning guidelines in a different costume. Answering it is a brief set of words that shapes the behaviour of an entire organization over a long period, which is two of Grove’s three sources at once. Not answering it was waffling, and waffling has negative leverage.
Neither of these is about how hard the work was. That is the whole point of separating activity from output. The pipeline rewrite was much harder.
Rank work by how many people it moves and for how long, not by how difficult or how technical it is.
Every later lesson that says “delegate that” or “your job is the system, not the change” is spending this idea. So is the warning in lesson 6 about new executives having too many opinions, which is meddling with the volume turned up by a title.
One caution before the next lesson. None of this says a CTO should never touch the technical work. It says touching it produces one unit of output. There are still reasons to do it, and lesson 4 is about the one reason that survives this argument.
Source: Andrew Grove, High Output Management, Ch. 3 'Managerial Leverage'
Answer to reveal the explanation. Nothing is scored.
1Asked what her output is, a director answers: 'architecture decisions, unblocking teams, and hiring.' What is wrong with the answer?
This is the exact answer Grove got when he asked a group of middle managers, and his response was that these describe what managers do while trying to create a result. A surgeon's output is a recovered patient, not the scrubbing and suturing. Naming activities as output is how a manager ends up optimising for looking busy at things that move nobody.
2You can write the year's planning guidelines in advance, or answer managers' questions as they hit them. Grove's model says these differ in what?
Grove's example is Robin, an Intel finance manager. Defining in advance what gets gathered at each stage, and who owns what, eliminates confusion for a large population over an extended period. Scrambling later to help one manager define milestones is the same activity with the multiplier stripped off. Timeliness is part of leverage, not separate from it.
3A CTO delays a decision on a migration for three weeks while gathering more input. What does Grove call this?
Grove is blunt that the lack of a decision is the same as a negative decision. Work can stop for a whole organization while it waits. He pairs waffling with a depressed manager as the two hardest kinds of negative leverage to counter, because their effect is pervasive and elusive: unlike a bad training session, there is nothing specific to redo.
4Which of these is NOT one of Grove's three sources of high leverage?
Doing the hard work yourself produces one unit of output and no multiplier. It can still be correct occasionally, for reasons lesson 4 covers, but it is not on this list. The three real sources are all about a single action reaching many people, or reaching one person for a long time.
5A senior manager sees a bad metric and hands the responsible engineer a detailed list of actions. Grove's term and verdict?
Meddling comes from a supervisor exploiting superior work knowledge, real or imagined, to assume command rather than letting the person work it through. The cost is not the one intervention, it is what the subordinate learns about the size of their job. This is the mechanism behind the new-executive anti-patterns in lesson 6.