The action: Ask for everyone’s 50% time estimate, pool up the safety buffer at the end, and keep everyone informed of when their turn is approaching for better projects.

The long form: Most projects run over time or budget. Eliyahu Goldratt in the book Critical Chain suggests a number of reasons:

1. No one knows how long a task will take. And since we don’t like delivering late, we provide a padded estimate, giving us perhaps 95% chance of delivering on time.
2. A task fills the allotted time, so once we have a generous estimate, we often spend more time on the task. Or we do more things in parallel (multi-tasking). This task-switching also slows the project down.
3. There is often no connection between the tasks in a project. Someone will dump their piece of work on our table, we will get to it when we have time, and then dump it on someone else’s table. But in the waiting periods, no value is added to the project.

Instead he suggests the following:

1. Ask for everyone’s estimate of how long their task will take with only a 50% probability. The stricter deadline helps people focus on the task and avoid multi-tasking.
2. Instead of each task having a safety buffer of time, pool that buffer at the end of the project. This pooling reduces the total buffer size, since there is less variability when combining individual variability. Say the sum of each person’s 95%-buffer was 30 days. If the sum of the 50%-buffer is 12 days you can keep 8 days buffer for the entire project – estimating the project at a total of 20 days.
3. Everyone keeps the next step of the process informed on progress, so that when the baton is passed to them they can drop everything else and complete their part quickly. “We expect to be finished with checking the legal requirements in two days. Be ready to start your part of the project on Thursday”.
4. If a step is completed ahead of time, the next step also starts ahead of time. Since everyone is informed of the progress they are ready to pick up the work once it is passed to them.

If you want to complete the project quickly, think of the progress along the critical path as a hot potato. When it is your turn to complete your part, pick it up immediately – not waiting until your estimated deadline – get it done and pass it to the next person as quickly as possible (whom you have kept updated so they are prepared to take it over).

What does this mean for you as a project manager? Your hardest task is probably making people comfortable with reducing their safety buffer. A 50% probability of completing their task in time means that half your your team will not complete their tasks on time. And that is fine! Make a promise you won’t hold anyone to it when they overrun. You do after all have a safety buffer for the entire project.

During the project your next priority is to get people to talk to each other, especially the people with tasks before and after them in the chain. You can also help in keeping an overview of where the baton is currently, whether there are risks or obstacles that you can help clear up. A whiteboard and regular (quick) meeting structure could help, or just a daily email to say “We are currently at step 2. Estimate handover to step 3 in two days.”