Notes on High Output Management by Andrew S. Grove

I finally got around to reading the classic High Output Management by Andy Grove, Intel's former CEO. When I saw it was available as an audiobook and after hearing about it from so many people within a short period of time (including in The Great CEO Within), I immediately jumped on the opportunity to read it.

I now understand why it's such a classic. This is the book that defined so many of the "best practices" that are taken for granted at top performing companies (OKRs, 1:1s, etc.). It was really helpful to hear the original/founding definitions of many important concepts like managerial leverage and task relevant maturity. I also learned a lot about how Grove recommended handling training (and why it's so important for managers to do), performance reviews, interviewing, running effective meetings. I also loved the idea of improving management like optimizing a factory.

Below are some of my notes and takeaways. I highly recommend this book to anyone leading a team or who wants to understand the reasoning behind so many of the management best practices in tech companies today.

---

​Introduction
Output oriented approach to management
The output of a manager is the output of their teams
Managerial leverage: high managerial output is based on doing tasks that have high leverage
How to inspire peak performance of individuals
Sports analogy 
Task-relevant feedback to adapt to change 
Plan the way a fire department plans: need to have trained team that can respond to whatever comes up
A responsive company should have fewer levels of managers
More direct reports per manager 
1:1 meetings
Teaching and exchange of info
OK if less frequent because status updates can be exchanged electronically
Managing your own career
You are in a business with one employee (yourself)
Need to keep improving
Your output is the output of your org
Learn to add more value

Foreword
Proper system design
Andy’s background
How to be great
How to conduct 1:1s
Training is manager’s job

Part 1 the breakfast factory

1 the basics of production
The three minute egg
Limiting or longest step of production
Plan production around critical limiting step (longest or most expensive)
Total throughput time
Time offsets and staggering steps based on individual throughput times
Production operations
Process production
Assembly
Test: unit and system
Rework
Limiting capacity of tools
Inventory and capital equipment and manpower trade-offs
Cost optimization
Better to choose in process tests rather than wait to end or choose some that destroy product
Incoming or receiving inspection
Have enough raw material on hand to cover consumption for the period of time it takes to replace the raw material
Adding value
Finished product more valuable than constituent parts
Fix problems at lowest value stage possible (at phone interview, at initial design, etc.)

2 managing the breakfast factory 
Indicators as a key tool
Sales forecast for day and variance from before for confidence
Raw material inventory
Condition of equipment
Amount of manpower
Quality indicator. Complaint log. 
Guard against overreacting so pair with other indicators for opposite goals
Inventory levels and amount of shortages
Joint monitoring
Quality and completion time
Any measurement is better than none
Measure output not just activity
What you measure should be physical and countable 
Pairing quantity measures with quality
The black box
Inputs and outputs with process as black box
Cut windows into black box to peek inside
Monitor credible leading indicators 
Linearity indicator of output over time
Can flash early warning
Trend indicators of output versus previous time or measurement against standard
Staggered chart of forecast in coming months to compare against previous forecast and reality
Controlling future output
Building to order versus faster turnaround
Build to forecast
Keep people responsible for performing against own predictions
Build in slack with inventory
Keep inventory at the lowest value stage like raw materials
Assuring quality
Inspection points
Reject defective material at lowest value stage possible. Reject before investing further value. 
Incoming or Receiving inspection
In process inspection during build
Outgoing or final inspection 
Trade off between checking quality thoroughly and not disrupting manufacturing flow
Deep gate like inspection holding the lien not moving until result vs monitoring inspection to just sample and later reject if sample bad but not holding up the line
Lean toward monitoring
Variable randomized inspections
Productivity
Output divided by the labor required to generate it
Change how fast we work or nature of what we do
Maximize Leverage
Work simplification
Draw flow diagram of all steps 
Count number of steps
Set target of reduction
Find ways to reduce steps

Part 2 management is a team game

3 managerial leverage
Output of a manager is output of their org and neighboring and subordinate units
Work is never done
Reports as medium of self discipline for author
Planning process is useful for itself and less for the result
Management by walking around
Ad hoc mini transactions
Information gathering, decision making, information giving
Nudging
Be role model and leader
Leverage of activities
Speed up activities
Do high leverage activities
Depressed and waffling manager are negative leverage
Delegation without follow through is abdication
Monitoring is not meddling
Delegate the activities you know best
Monitor at lowest value stage of project like rough draft
Variable sampling scheme for monitoring frequency l depending on the subordinate’s task specific maturity
Go into details only randomly 
Monitor subordinate’s decision making process
Figure out how good the thinking is
Increasing managerial activity rate by increasing the speed of the line
Time management techniques
Handle a paper only once
Standup meetings for brevity
Turn back to door
Figure out what is immovable and plan other things around it
Batching similar tasks
Factory versus job shop
Factory run by forecast and planning 
Calendar is medium of forecast
Don’t just use calendar to respond to others requests
Take active control of calendar and fill empty spots with important low urgency work
Say no to extra work beyond capacity to handle
Say no earlier rather than later
Time is your one finite resource
Allow slack
A bit of looseness in schedule
Carry raw material inventory of projects
Can be done in free time and no deadline
Things to improve org over time
Number of subordinates to have
6-8 ideal
Allocate a half day per week to each subordinate
Interruptions
Strive toward regularity
Smoothe out our workload
Prepare standard responses
Can delegate to others
Batch responses
Batch questions for 1:1s
Use of indicators to speed up responses
Hold open office hour after period of individual work

4 meetings: the medium of managerial work
Critical way to supply info and know how
Mission oriented meetings ad hoc to make decision
Process meetings should have regularity
1:1, staff meeting, operation review
1:1s
Mutual teaching and exchange of info
Frequency depends on task relevant maturity of that subordinate
Weekly if inexperienced
Hour long duration
Do meeting in subordinate work area to see how he works
Subordinate sets the agenda and see it as his meeting
Content
Performance indicators
Problems and updates since last meeting and upcoming plans
Potential problems
Issues that nag the subordinate
Didactic management principle: ask one more question to get subordinate to talk more
Role as coach
Both should have copy of agenda and both take notes on it during the meeting in outline form to think and categorize
Take notes to follow up next time and show commitment
Create Hold file to keep actions and ideas for future follow up
Talk about intangibles like doubts and job satisfaction
Schedule on a rolling basis as opposed to standing meeting to avoid conflicts like due to vacation
High leverage activity
1:1s useful at home too. Parent takes kid to restaurant 
Staff meeting 
Manager and all subordinates
Discuss anything that affects more than 2 people
Follow agenda and have open session at end
Supervisor as moderator and facilitator
Operation reviews
Formal presentations to peers in other teams
Pace presenters with gestures
Mission oriented meeting
Held ad hoc to produce outcome or decision
One person as chairman who calls meeting must know objective
Don’t hold meetings unless justified because will cost $100 per hour per attendee
Don’t waste time
No more than 8 attendees
Don’t let others delay conversation 
Send out agenda ahead of time and roles others need to play
Send out minutes after with all action items

5 decisions decisions
Knowledge and position power
Ideal model
Free discussion
Reaching clear decision
Everyone gives full support
Work out decisions at lowest competence level
Peer group syndrome
Peers will be afraid to push for one idea
Peers look to more senior manager to help guide decision and process
Just pick someone who has most at stake to be in charge
Striving for the output
Don’t push for a decision too early
Don’t let near consensus drift on forever
Settle 6 questions in advance
What decision to make
Who will make it
When to make
Who to consult
Who to ratify or veto
Who to be informed

6 planning
Forecast market demand
Consider factory output compared to demand and adjust as needed
Environmental need
Consider needs on your group from your environment
What will customers expect of you now and later
Present status
Examine projects in pipeline
What to do to close the gap
And what can you do to close the gap
Management by objectives
Where do I want to go
How do I know if I’m on the right track (key results)
If plan yearly, then need to look at key results quarterly or monthly
Need focus. Pick a small number of objectives and key results
Nesting hierarchy of objectives
Just one input into performance review
Very specific wording and dates
If manager meets his objectives then supervisor meets his

Part 3 team of teams 7

7 breakfast factory goes National 
Centralization vs decentralization trade off when franchising
Some functions central, some regional
Regional distribution centers
Management is a team game
Team of teams

8 hybrid organizations
Interconnected teams
Mission oriented or functional
Compromise between the two extremes
Responsiveness vs leverage
Mix of Business divisions and also functional orgs that support them as internal subcontractors 
Two thirds of employees in functional units at intel
Business division org can respond more quickly to market
Hybrid is best structure
Grove’s law: all big businesses end up in hybrid form
Central planning org isn’t the best final answer
Middle managers is the answer

9 dual reporting
Matrix management
Whom should security personnel report: both to corporate and to local
A person can have two bosses
Group decision making
Coordinated body of peers in addition to individual business divisions 
Two plane organization 
Engineering managers counsel
Name on two org charts which are orthogonal
Transitory task force teams

10 modes of control
Three possibilities: free market forces, contractual obligations, cultural values
Management spells out values and shows by example
Nature of person’s motivation 
Complexity of environment and motivation of person affect choice of mode of control
Individual self or group interest versus high or low CUA (complexity factor)
Expectations can be as binding as contractual obligations

Part 4 the players

11 sports analogy
Management is a team activity 
If person not doing the job, then either is not capable or not motivated
Test: if their life depended on it, could they do it?
Most important management job is to elicit peak performance from subordinates
Two levers: training and motivation
Performance, not attitude, is desired result
Maslow Needs, drives, motivation
Physiological needs
Security and safety needs
Social affiliation needs
Esteem and recognition needs
Once need satisfied, it’s self limiting 
Self actualization needs; continues to motivate
Competence driven or achievement driven
Set objectives just out of immediate grasp to push yourself to achieve 
Money and task relevant feedback
Money as objective vs measure of achievement which is not self limiting
Fear of failure
Building indicators to allow competition between divisions to encourage harder work
Manager as coach, critical of team and pushing to improve

12 task relevant maturity (TRM)
Manager job to elicit best performance from employees
No optimal management style
Difference from a person’s general competence
Start with more structured management for low TRM and make less structured with high TRM
Delegating vs abdicating
Give instructions vs establish objectives
Like parenting changing over the years

13 performance appraisal
Task relevant feedback
Assessing performance 
Clearly define what you want from person
Judge performance of manager and the group under them
Assess performance not potential. Substance over form
Delivering the review
Level
Listen
Leave yourself out
Target just a few key areas
Shouldn’t be any surprises if been having good 1:1s
The blast
Better to give written review in advance to employee then discuss face to face

14 two difficult tasks
Interviewing
Candidate should do most of the talking
Interrupt when going off track 
I’d like to change the subject to xyz
Convincing someone not to leave 
Need to deal with situation right at first mention
Let them talk it out
Ask more questions
Don’t argue or lecture
Convey why he’s important
Ask for more time and to speak with supervisor
Try to keep with firm including help transfer to another division

15 compensation as task relevant feedback
If absolute amount of raise matters, that person is after physiological needs
If relative amount matters more, then they are after self actualization
Performance bonuses 10-50%
Team and individual role in bonus
Subjective or objective elements
Multifactorial model
Peter principle: people promoted to level of incompetence

16 why training is the boss’s job
Manager should do the training himself 
High leverage activity
Canned courses from outside don’t help and not specific to how the company works
Training should be a process not an event
University catalog of internal courses

One more thing (homework to do)
Production
Leverage