In my previous post on making project delivery promises, I wrote about the need to set delivery dates based on a clear understanding of capacity and workload. During project execution, the promises don’t shift, but reality does. For project teams, the need to assess the risk of missing the delivery date is imperative to the success of the project, often more than budget or scope. We call it a “risk”, because there is always uncertainty in predicting the completion date. Murphy lives, but we don’t know when he’ll visit – or if he will bring his friends!
Traditionally, managers have tried to reduce this uncertainty by projecting completion times for projects based on the sum of individual task durations on the critical path. However, these task duration times are always suspect because there is uncertainty in the prediction. When managers ask for task duration estimates, resources and teams typically add “safety” time that anticipates setbacks (setting expectations to not be late).
In themselves, these safeties are no big deal (in fact, the need for safety time is essential), but when added together, the collective times make the predicted project duration much longer than it needs to be, clouding the actual schedule risk and adding unnecessary delays to project completion.
What is a project manager to do when dates must be met and there is this uncertainty on top of uncertainty?
I can’t go into the full solution in this post, but you can learn about probabilistic project planning in our white paper here. The general solution to the schedule risk management problem is to gather estimates of the most likely task durations without safety from each task team, breaking apart the work from the safety. (These estimates don’t need to be perfect to be useful.) Then, create a time buffer for the entire project, approximately equal to half the critical chain.
During project execution, task completion and buffer consumption are tracked collectively on a regular basis; the project team is able to make an accurate assessment of schedule risk and rapidly respond appropriately. Distributing the buffer to various tasks as needed, applying it to those task challenges most likely to threaten the overall project. We call this process of review and allocation “buffer management”.
Remote teams require special considerations to make the best of using buffer management.
- Distance insulates leadership from the problems encountered throughout the project, impeding their ability to assess real progress
- Without a timely way to communicate and share true progress, obstacles cannot be identified in time to respond without jeopardizing the project delivery date
- Distance obscures the actual work required and the uncertainty of on time completion of any task in the project
- Different teams have different ideas about risk and work, further obscuring the work picture
- Local variations create confusion about what is or should be reported, impeding clarity
During a project to install a critical repaired component off the coast of Africa, one of our clients in Houston was able to coordinate distant suppliers, various engineering groups, project managers at the local site, and the installation crew on another continent.
They compensated for the leadership insulation obstacle by sharing a common view of the project status on a regular conference call and screen sharing software. Everyone had the same plan and rules of engagement.
The timeliness obstacle was overcome by tying communication to the tempo of the project. Rather than being a slave to a weekly or daily huddle, the team was flexible according to the speed at which work was completed and the perceived risk. Early the project, the huddles were weekly. As the project neared completion, the frequency of huddles increased – during installation, they were twice per day.
The work identification obstacle was overcome by again, being flexible and responsive. The team was constantly aware of the amount of safety versus work in the critical chain, so any changes in either dimension generated an immediate response – either to add resources or work process revisions in the case of increasing risk, or changing focus in the case of decreasing risk.
Local variations and interpretations of the work were encountered – however, a systematic review with and some training of the remote teams early brought everyone together. When it was time to execute, the entire team had the same goal and understanding of the critical assumptions underlying the execution plan.
The result was a very successful repair and a very happy customer. They were able install the new components during the time window specified, even when that window was moved early during project execution. Not only that, they compensated for a doubling of work scope and completed the repair in
25% less time!
The keys to overcoming the remote collaboration obstacles to managing schedule risk are straightforward:
- Have a common understanding of the work plan – make it visual!
- Clearly communicate the understood schedule risk
- Be prepared for Murphy. He will be there.
- Be flexible in timing
Read more about how to make realistic risk assessments with remote teams in our new eBook, Remote Control: How to Get Productive Teamwork from Distributed Project Teams.