Leadership Lessons on Project Management: What Project Sponsors Can Learn from Swiss Cheese

 

Leadership Lessons on Project Management: What Project Sponsors Can Learn from Swiss Cheese

Professional techniques for project management have made a huge impact on the success rates of projects in the past three decades, but there’s no denying that we still have a long road ahead of us. Statistical information from G2.com gives us a peek into this challenge. Most organizations have a 70 percent project failure rate, only 28 percent of companies use project performance techniques, on average one in six projects has a budget over-run greater than 200 percent, and only 2.5 percent of companies complete 100 percent of their projects successfully.

That’s probably not a surprise to you, since I’m preaching to the choir here. The question is what else we can do to improve project management reliability. I believe we can learn much more from a technique called the Swiss Cheese Model of Accident Causation which has been used to dramatic effect in several industries including aviation, to improve reliability.

Accident Prevention and Swiss Cheese

The improvement in aviation reliability from the early days of World War 1 where pilots flew “on a wing and a prayer”, to 99.999999% reliability today has been an inspiration to several industries, including medical safety and computer security. Flight safety in the early days of aviation was not a given. The airplanes which were made of fabric, glue and wood for the most part made flying more a game of skill than process. Takeoff and landing crashes were common. The intervening decades have seen an improvement in both design and methodology. The Swiss Cheese Model was one of the risk minimization techniques employed.

The Swiss Cheese model of accident causation states that human systems are similar to slices of swiss cheese that are placed vertically in front of each other. The holes in the cheese represent defects or weaknesses in each system and tend to be of different sizes and positions. If a line can pass through the stack of cheese holes then that represents failure of the system as a whole, leading to an accident. The goal in designing the system is to reduce the probability of an accident by improvements on both “holes” and “number of slices”.

James Reason, who created the Swiss Cheese Model (SCM), believed that accidents can happen for four reasons i.e. organizational influences, unsafe supervision, preconditions for unsafe acts, and the unsafe acts themselves.

Organization-centric vs. Individual-centric Causes of Project Failure

Let’s try and get this back to project management. There’s something intriguing about the number 70, which indicates the percentage of failures in project management. That number shows up in other industries too. In an interesting analysis by the Colorado Firecamp, called the Human Factors Analysis and Classification System, it is pointed out about 70 percent of aviation accidents can be partly attributed to human error (Shappell & Wiegmann, 1996). However, we know the attribution of a large percentage of failure to pilots alone is misleading. Their analysis rightly argues that to off-handedly attribute accidents solely to aircrew error is like telling patients they are simply “sick” without examining the underlying causes or further defining the illness. In other words, we need to look at not just the final visible cause (individual-centric), but also to the previous layers of issues (organizational-centric). It’s well known that aviation accidents are the end result of a number of causes, only the last of which are the unsafe acts of the aircrew (Reason, 1990; Shappell & Wiegmann, 1997a; Heinrich, Peterson, & Roos, 1980; Bird, 1974). That’s true in project management failures as well. Attributing most of the failures to the project manager and the project team ignores the three other preceding layers in the swiss cheese model i.e. organizational influences, unsafe supervision, and the preconditions for unsafe acts. That’s where project sponsors need to play an important role.

The Sponsor’s Role in Addressing Organizational Influences, Unsafe Supervision, and Preconditions for Unsafe Acts

Unsafe acts in Project Management are generally based on the foundation laid by Organizational Influences, Unsafe Supervision, and Preconditions for Unsafe Acts. So, for example, choosing a weak software solution in a given project is an unsafe act.  However, when we step back a layer it’s possible that the poor choice was based on lax procedures in either technology standards, or in matching-and-mapping process design to software solutions (precondition for the unsafe act). Step back another layer and we may find that the project board did not ask the right questions while overseeing the project (unsafe supervision). Go back yet another layer, and we might find that the siloes between the IT architecture group, the software procurement function and the project management group sub-optimized the software choice while trying to optimize individual department results (organizational influences). Who needs to help the project team across these layers?

I believe project sponsors play a key role here. To be clear, it is impossible for any given project sponsor to systemically fix all issues across all four layers. A good sponsor, however, is educated enough in these practices to work alongside the project team and guide them in avoiding these traps when they cannot be systemically addressed.

In closing

As Vice-President at Procter & Gamble’s famed Global Business Services and IT, I considered the ideal project sponsor to be a person who had four qualities. They had to own the organization that needed the project, they needed to have enough time to devote to the team, they had to have enough knowledge of the project’s subject matter to add value, and they had to be effective in breaking organizational barriers for the project team. This worked extremely well for us. Unsurprisingly, these are also the qualities that help project sponsors address all four layers of the Swiss Cheese model.