How big things get done published the following breakdown:
- 100% of the mega-projects analyzed
- 47.9% finished on budget
- 8.5% finished on bugdet and time
- 0.5% finished on budget, time, and produced the expected benefits
Later, the book analyzes what are the prerequisites to get into the 0.5% Olympus:
- Know why all the time: is this code rewrite benefitting the product?
- Think slow, act fast:
- Make mistakes in the planning phase when screw-ups are cheap
- The more time the execution phase takes, the more you expose yourself to black swans (ie, catastrophic and unexpected events)
- Mitigate risks:
- Spike code to understand if and how something can be done
- Ask people who worked on similar features what went wrong/could have been done better
- Estimate using anchors:
- Hofstadter’s Law: “It always takes longer than you expect, even when you take into account Hofstadter’s Law.”
- We always estimate with the happy case in mind and, even the most paranoid, cannot prepare for unknown unknowns
- Ask others what was their experience building a similar thing and use their numbers as a starting point
- Lean on experience:
- Involve people who've worked on similar features in the past
- Learn from other similar projects—and no, you are not working on anything so unique that cannot draw from experience
- Use boring tech: "remove the words custom and bespoke from your vocabulary"
- Build a team
- Iterate:
- "we're terrible at getting things right the first time"
- "I have not failed ten thousand times," Thomas Edison said. "I’ve successfully found ten thousand ways that will not work."
- "If something works, you keep it in the plan. If it doesn’t, you “fail fast,"
- When talking about the success of the Empire State Building, "workers didn’t build one 102-story building, they built 102 one-story buildings."
The following sentence made me stop and think about my fuck-ups:
When delivery fails, efforts to figure out why tend to focus exclusively on delivery. That’s understandable, but it’s a mistake, because the root cause of why delivery fails often lies outside delivery, in forecasting, years before delivery was even begun.