Seventy years later, 0.5% of them are.
In the late 1950s, Morgan Walker (DuPont) and James Kelley (Remington Rand) built an algorithm to identify the longest chain of dependent tasks in a project. They first ran it on DuPont chemical plant shutdowns in 1957. The method saved 25% on shutdown time. They published the founding paper in 1959.
It is now used on every megaproject in the world.
It does not seem to help.
In 1957, Sydney commissioned an opera house from Danish architect Jørn Utzon. The plan was clear:
- Budget: $7 million
- Timeline: 4 years
- Opening: Australia Day, 26 January 1963
It opened on 20 October 1973. The final cost was $102 million. Ten years late. 1,357% over budget. The architect quit before completion. The state lottery paid for most of it.
In How Big Things Get Done (Penguin Random House, 2023), Oxford's Bent Flyvbjerg analysed 16,000 large projects across 136 countries spanning 90 years.
48% finish on budget.
8.5% finish on budget and on time.
0.5% finish on budget, on time, and deliver the promised benefits.
CPM is the most widely taught project management technique in business schools. The technique works. The discipline of using it does not.
The critical path is not your forecast. It is your test of attention.
Where in your current plan are you treating estimates as schedules?
