Projects fail all the time. Politics are usually blamed, but that’s rarely the case. Usually it’s sociology. Our work is more sociological than technical.
Even leaders who concede that the challenge is people seldom manage that way. People not tech.
Development != production. In production, errors suck. In development, they’re quite useful
Team dynamics are paramount. Catalysts can be worth 2 workers
Leave time to think, learn, brainstorm, etc. You can’t always be in 100% development mode.
Workaholics and death marches are futile
Quality is a means to higher productivity
Parkinson’s Law doesn’t apply, but certainly don’t treat people like it does.
Companies can exhibit Parkinsonian behavior though - busywork can run rampant at larger organizations.
Tech and management is full of “lose weight while you sleep” kinda things. Don’t believe the hype.
There are a million ways to lose work
(page 46) More myths that don’t really solve things: programming language, # years of experience, # of defects, salary
Things that do matter: teammates. Good people cluster from org to org, and within an org. Some companies outperform others 10:1
If you hear “you can’t get anything done around here from 9 to 5” that’s bad. Take action.
Workspace quality and product quality are directly related.
If people hide out to get work done, your workspace sucks.
(page 79) Music will stifle creativity and ah-ha! moments
Vital space: sharing space when on same project is beneficial
Master plan vs. meta plan: master = suffocating design, meta lays out growth rules
Idea: a Melrose Place type of office would be awesome
Let folks establish their office design
You need windows. Nobody would think of living in a house without them, no hotel or apartment would dare have rooms without them. Why do office workers tolerate this? Window politics are a crock of shit.
Indoor space as well as outdoor space is necessary
Intimacy gradient: one enters into the public space, then moves to the workgroup space, then to the private areas that employees consider their sanctuary.
Communal lunches are a huge boon to teams
Alexander’s patterns work because they’re in harmony with humanity and man’s fundamental need for individuality and community
Get the right people, make them happy, turn them loose
Require a portfolio & explanation from interviewees
Aptitude tests are good for self-assessment, but not for hiring. They tell where you may grow within an org, but not necessarily whether you’ll be a good initial fit.
Have interviewees present on something germane to the company and job to the people in charge of hiring them.
Companies that are all about “the best” are best, and filled with people who came up through the ranks
Deterministic vs. Non-deterministic systems: if a system has a high degree of “ad-hoc-ness”, don’t try to automate it. It’ll just be forever in maintenance mode.
Voluminous documentation is a problem, not a solution.
Better to get people on the same page: training, tools, peer review
Hawthorne Effect: change = boost in productivity
Teams are one thing, jell is a whole ‘nother.
A common goal is essential to jelling.
Goal alignment vs attainment
Jelled teams: SWAT attitude, joint ownership, shared space, project artifacts are decorations to the workspace
Jelled vs. Clique = breeze vs draft.
Teams are grown rather than built
- Be defensive and don’t trust team members. Don’t let them make mistakes
- Add lots of process and bureaucracy
- Cubes and physical separation
- Fragmented time & members on multiple projects
- Reduce quality
- Crazy/phony deadlines
- Clique control
“Most organizations don’t set out to kill teams, they just act that way”
Spaghetti dinner: provide opportunities for the team to jell.
Look for excuses to get them offsite.
Awards for insubordination
Look for a shared sense of something (humor, music, philosophy, etc)
Natural authority: journeymen don’t mind listening to the master craftsman
Cult of quality
Closure: milestones, confirmations of being on target
Elite team: formed identity, they pull toward a common goal whether management likes it or not, sometimes arrogant and haughty sometimes - accept this
Don’t break up the Yankees
Good team structure is a network, not a hierarchy
Everyone doesn’t have to be the same. Junior guys and others can help and teams can rally around making sure they’re brought up and integrated into the fold.
Inject some chaos back into the work. The trend over a company’s life is to rein in the chaos, but it’s actually fun & rewarding for teams to confront and rein it in themselves.
(page 263) Examples:
- pilot projects (be discreet about what you’re piloting though - don’t pilot 10 things at once)
- war games: compete on something needed by the org., use winner(s) in production & track bugs, etc.
- brainstorming: analogy, inversion, immersion
- provocative training, trips, conferences, celebrations: these are super valuable early on in team formation. If possible, get them traveling and dining together.
Loosen up job descriptions and let members roam a bit. Intrapraneurs are very valuable.
Revisions to First Edition
Beware of over competing
Other teamicidal things:
- salary and merit reviews
- singling out workers for praise
- awards and bonuses tied to performance
- performance measurement
Actions that treat team members differentially are teamicidal
Sports teams are a bad metaphor. Musical ensembles are more apt - if one’s great, and everyone else sucks, it’s a failure. Win or fail together as a unit.
Process isn’t worth a shit if the product sucks.
Resistance to change is automatic and the reaction is emotional, not logical.
Remind members that “All improvement involves change”
Make people feel safe with change: accept hiccups and don’t demean people for mistakes
Experience != learning
Learning is limited by turnover. The greatest learning organization doesn’t mean shit if turnover is high.
The ability for a company to learn and redesign organizationally is a serious advantage.
Usually middle management is where the learning core is, but communication has to be open and serve a common purpose. If leaders are looking out for #1, learning is stifled.
The ultimate management sin is wasting people’s time.
Regular get-togethers are likely serving a ceremonial purpose to stroke a manager’s ego and reinforce his bossness. Save ceremonial meetings for milestones, launches, etc. Orgs and people need ceremonial activity, just don’t make it meetings.
Early overstaffing: adding too many people early to a project can make it a clusterfuck. Take people on later, after planning and early design.
Time fragmentation and gear switching are huge time drains. You can’t really measure the time lost to get back into the flow of one task to another.
We have a strong need for community and the workplace is usually a surrogate for the town. The U.S. hasn’t had a strong sense of town since the 20’s (the car).
There’s no formula for how to build a thriving community at work, but it’s one of the loftier goals a person can pursue.