“High Output Management” Takeaways

Coming into reading this book, I’ve been challenged in how I think about management and hierarchy. I found it to be inherently rude. I think because it automatically restricted autonomy. But what I’m accepting is that restriction of autonomy, if agreed upon, is a necessity for even small things to get done. Whether it’s a company or a social movement, separation of responsibilities is necessary.

Just think about the 1960s bus boycott, and the coordination amongst funeral homes and neighbors to set up an alternative commuting network. People had to do routes, get fuel, do basics of communication, etc. because the boycott needed to sustain. And for all parts to work in tandem, there needed to be people gathering and distributing information. One name is manager, but maybe a word I like better is a conduit. It’s more about empowering from below, like the roots of a plant. I think this book supports that notion.

If a group is more effective without a conduit, then remove them.

  • people are nodes in an information network
  • when products are indistinguishable the deciding factor is time
  • have a tolerance for disorder but don’t accept it. Do your best to drive everything around you towards order

The basic:

  • every employee produces something
  • Output of a manager is measured by the output of the organizational units under his or her supervision or influence

Change in methods of information transfers (ex: slack) means less time needs to be spent info sharing and more can be focused on problems/concerns

No one owes you a career. It’s on you to improve and compete with millions

“You can’t be an optimist about the future until you have survived the crucible of change”

Leading vs. lagging indicators. A leading indicator could be an analytics tool set up in advance. A lagging indicator is when your bank account hits 0 and you just found out. Harder for people to accept and act on “leading bad indicators” when they’re optimists for the future because it’s often incorrectly lumped together with the nay-sayers


  • production
  • Managerial leverage (continuously choosing to do highest leverage tasks)
  • Athletic desire for peak performance

A manager output = output of her organization + the output of the neighboring organizations under his influence = L1xA1 + L2xA2

Knowledge is thus worthless unless it can be effectively transferred to the rest of the org. That’s what being a manager is about

Difference in knowledge-power people and position-power people. Need both to make decisions affecting the org years down the road

The overall shape of an operation is determined by limiting step (ex: the boiled egg, not toast)

Production operations = process, assembly, test

Indicators are a key tool

  • Sales expectations
  • Material inventory
  • Equipment condition
  • Manpower
  • Quality

Don’t steer where you measure. Need counter measurements. Ex) inventory paired with incidents of shortages. No incidents no need to worry

Useful indicators cover work unit output, not activity

It’s surprisingly hard to get people to have confidence in indicators. You have to work at believing their validity

Inventory should be kept at its lowest value stage (ex: a raw egg)

It’s easier and less costly to get rid of inventory at its lowest cost (does that work with features? Business models? Data before it enters a forecast?)

Gate inspection (allows low cost materials to pass through) vs. variable inspection (harder logistically for us habit heavy humans)

Visa processing and classification of different response offerings created overhead. Focus rather should be on speed improvement for all. Follow IRS lead and group based on predetermined criteria and do sampling tests.

Managers digging into problems are examples of variable inspection. If it was every operation checked the employee would feel their meddling and would not feel responsible for their work. Think of this method to prevent micro management

Productivity can be increased by higher rates of activity or by increasing leverage of activities

Leverage is obtained with automation but also “work simplification.” Write our all the steps of a workflow and see if you can get rid of 30-50%

It’s hard to distinguish output from activity in soft professions so it’s key to examine and stress output.

Soooooo measure output, measure output, measure output.

Managerial leverage

  • information gathering/giving
  • Decision making
  • Nudging
  • Role model

Casual verbal exchanges are the best provider of timely information, which is the most valuable. Besides being archival, written form is a process that self-disciplines for precision.

To avoid the awkwardness of walk through without a task, programmed visits are made and then ad-hoc mini transactions can occur

Meetings are a medium like memos

Batch activities

Subordinates should be 6-8 with half a day spent towards each. That’s if it’s a supervisory role. This prevents over involvement.

Calendars are the production lines of managerial work. Control the time with no’s to work beyond what can be handled and filling in blanks.

Bring regularity to the irregularity of interruptions. This turns the distractions of interruptions into a batched process pattern, that can be understood and respected by subordinates and later templated or delegated our once it has structure. It’s unacceptable to avoid problems.

Making regular what is irregular is a fundamental production principle

Meeting Types

  • Process oriented (info exchange)
  • Mission oriented (problem solving) => decision


  • No less than an hour or it confines to the basics
  • Agenda is set by subordinates because manager can’t spend time preparing for 8
  • Supervisor needs to facilitate to get at what’s bothering them. “Ask one more question!”

Process oriented staff meetings should ideally have an inactive supervisor and active participants. Staff are the ones working through the problem while supervisor keeps people on track.

Regularly scheduled process oriented meetings should ideally resolve everything before it’s mission oriented. In practice they catch 80%

Malorganizarion is not when you spend 25% of time in meetings, it’s 25% in ad-hoc mission oriented ones

Decision making process

  • FREE discussion (difference of opinion needs to be aired out. Ask why)
  • Clear decision (scope setting)
  • Full support (no change in scope and intervention)

The people who do this best are those who have comfort and confidence in the org values and those who are fresh from college where this environment in class is normal

Decisions must be made and communicated at the lowest component level. Ideally a expect agreement but it might take time to get there

Peer plus one - when a homogenous friendly group needs an outside force to shape the meeting. Also there is hesitation to disagree in a peer group.

Strive for meeting output. Don’t push for consensus but if all perspectives were raised and people start to drift from near consensus, reign it into a decision

For improved managerial output to occur and a decision needs to be made ask:

  • what decision needs to be made
  • When does it have to be made by
  • Who will decide
  • Who will need to be consulted prior to making the decision
  • Who will ratify or veto the decision
  • Who will need to be informed of the decision

This preplanned structure prevents politics. If “Johnny-comes-late” vetos, political maneuvering is inevitable. For the vetoing individual, if it’s drastically different give people time to adjust and get their heads together. Don’t walk away from the issue.

If you can make decisions without consulting anyone, so can everyone one else. (So don’t allow for that)

Breakfast Factory goes national

  • the constant tug between centralization and decentralization happens as all organizations become large.
  • when does something become the CEOs decision and when is it a local branch manager? When does autonomy allow for best adaptation and when does it allow public opinion of quality, etc. to be jeopardized? When is it best to have layers in between, like regional distribution?
  • the managers output is the output of the organization supervised by him, which is not just a team game, but requires fashioning a team of teams that mutually support one another.

Two forms of organizations (usually a mix)

  • mission
  • functional

“Good management rests on a reconciliation of centralization and decentralization” — Alfred Sloan (or might we say a balancing act to get the best combination of centralization and decentralization)

Functional groups are the supporting teams to mission oriented teams, who have to specialize a specific craft and adapt rapidly. At intel this means manufacturing is a functional group, so production doesn’t have to scale back and forth within functional units and knowledge managers can persist and be leveraged by each business unit as they rapidly adapt

There is only one time you need to organize portions of a company into a mission-oriented form. “It is that the individual units can stay in touch with the needs of their business or product areas and initiate changes rapidly when those needs change” that’s it.

  • All else is functional. But any business has to respond rapidly to changes, so it’s often that much of a business is organized into mission oriented units.
  • This hybrid org is not only inevitable to Grove, but there is no other alternative after a lot of experimentation at Intel.

Where a company falls on the spectrum of those two extremes will shift over time due to practical concerns. That shift has two concerns:

  • optimum and timely allocation of resources
  • efficient resolution of conflicts arising over that allocation

Matrix management and dual reporting

  • eventually managers find themselves rising to spaces through technical tracks and managing divisions they have no experience in.
  • you need immediate operational priorities coming from a functional supervisor but you also need them coming from a technical supervisory relationship
  • horizontal functional peer groups as a means of shared problem discovery and solving

Hybrid organizations are ambiguous like democracy. They’re not great in and of themselves but just happen to be the best way for any business to organize

  • lol
  • neither the form nor need for dual reporting is an excuse for needless busywork. Mercilessly slash away unnecessary bureaucratic hinderance
  • “We should not expect to escape from complexity by playing with reporting arrangements. Like it or not, hybrid organization is a fundamental phenomenon of organizational life”

Know how managers are seen as operating on two planes. They’re one report can be or someone at a specific mission oriented team but their other can be a functional knowledge sharing across all teams. That cross functional knowledge manager had more leverage

  • Planning bodies usually sit on separat planes as operating bodies
  • multiple plane are fine. They recognize strengths of individuals and often can have role reversal in terms of who answers to who. Grover answers to the chairman w

Three methods of control. Management decides which is most appropriate, because full selflessness isn’t always best:

  • free market forces
  • contractual obligations
  • cultural values

Interplay of Individual Motivation and Complexity, Uncertainty, and Ambiguity (CUA)

  • self interest + low CUA = free market
  • group interest + low CUA = contractual
  • group interest + high CUA = cultural
  • self interest + high CUA = CHAOS. Nothing works.