Notes on "Peopleware" by "Tom DeMarco"

Human resources

• Projects fails to due sociological reason, not due to technical reason

• Knowledge base work is fondamentaly different from production work in that :

  • People enjoy their work

  • Pressure on people will make them work faster, not better. In the long run it's an issue

  • You should allow a buffer for errors, people learn from errors that whats makes them better in the long run.

• Overtime is always followed by an equivalent or more important downtime. People need to catch up with their life at a moment or another.

• Don't reason in a fixed 40 hours work week. Since we are in a knowledge base work, people might work or think about work OUTSIDE of work.

• If you push for quality, (not market quality but worker quality) people are more motivated to work harder.

• The authors give examples of the Japanese industry and 80's Hewlett-Packard. We could also think about the Blizzard slogan during the 20's "Release it when it's ready" Reddit

• If you are commited to quality, you give the power to the worker to veto if a feature hasn't meet the standard for quality

• According the the author, the Parkinson law that state "Work complicates to fill the available time" isn't verfied by data.

• According to the book, the best productivity is achived when the project is not estimated

• The authors conclude this part with "The manager function is not to make people work, but to make it possible for people to work". In the regard the classical role of "Scrumaster" fits to that definition

The office Environment

• People might work better with natural light and low noise level

• By optimising space inside workspace, you might be losing efficiency of worker. The ratio is around 20:1

• $$Ratio\ = \frac{ worker\ efficiency}{ space\ saving} = 20:1$$

• Don't interupt people, your losing in efficiency.

• Their might be a ratio in productivity of 10:1. Between the best developers/organisation and the worst

• People work best when they enter the "Flow" it's a state of deep focus where people lose the sens of time. You need calm and concentration to enter that state of mind

• $$E\ factor\ = \frac{ Uninterrupted\ Hours}{ Body\ Present\ Hours}$$

• If you can use a Asynchrone way of communicating it might be more efficent call 😈 VS mail 😇

• The author state that it's more complicated than that, relying only on asynchonous communication and mail could be harmful. From my experience, relying only on asynchronous communication canal, is harmful especially during long period of remote working (Covid 😷)

• Having a door is nice

The right people

• Get the right people, make them happy so they don't want to leave, turn them loose

• Leadership is a service to people. You don't lead people from behind, you show them the way.

• There is two kind of leader, the one who talks and the ones who do.

• When you hire someone don't forget to test the candidate capability to communicate. He should also integrate well well with the team. Doing an "audition" which is a little presentation about a past work is a good idea according to the author

• No one whould hire a jungler without having him jungle first, the same should apply to software developers. The author recommend that you check the portfolio of the candidate.

• Diversity in a team change the way the team works for the better (women, socio-eco, culture). A team need time to jell together. If you can don't hire too fast

• The new technology of a generation becomes the environment of the next one

• If peaople are not focus on a meeting it's a sign of an unhealthy meeting culture (no laptop + standup meeting)

• A company should be able to answer "What's your turnover rate ?"

• A healthy should act has it is the best, and build a culture around it. Being the best requires committement on both sides (company & employees)

• P&L do not take into account human investement in knoweldge base organisation

Growing Productive Teams

• You should look for a "jelled" teams. A "jelled" team is a team comitted to achieve the goal they are responsible for. Teams members are like "glue" and will do whatever they can to achieve things. They purse goal with an enormous energy

• Most of the time no one cares about company internal KPI it's not what inspire people

• "Jelled" team just feels happy

• How to kill a team :

  • Defensive management
  • Bureaucracy
  • Physical separation
  • Fragmentation of people's time
  • Quality reduction of product
  • Phony deadlines
  • Clique control
  • False promotion of "best place to work"
  • Heterogenous Overtime

• Competition shouldn't be promoted. It is inhibiting team work

Fertile soil

• A corporation is a non deterministic processus. Don't rely on methodology to do your work.

• If their is no risk in a project to be done, it's not an interresting work

• No laptop in meeting, a meeting should have an agenda. Call for a meeting when a decision must be made.

• Bring has few people has possible into a meeting. Less is more

• Avoid ceremonies, meeting should be ad-hoc (not planned)

• Life is short you don't need to know everything in order to do something

• There shouldn't be "silence give consent" implicit rule

• Your goal should be to build a community

It's supposed to be fun to work here

• Idea to make work fun : Pilot projects, hackathon, brainstorming, training, trips, conferences, celebrations, etc..

• You should enjoy your work and work with people that enjoy it too