Finding Lost Time in Software Development

"The only thing that is constant is change." This famous quote originated from a Greek philosopher named Heraclitus. For many this statement still rings true. In software development, this notion is simultaneously its biggest asset and weakness. Another area that permeates teams with the same voracity is Parkinson's Law. This adage states "work expands so as to fill the time available for is completion." It was coined back in 1955 by Cyril Parkinson in The Economist. Everyone has had first hand experience with this concept. Starting a day with high expectations of productivity and losing it in the warm of a summer afternoon. Although this phrase was born before the modern era of computing, it has found a home in software development. Similar to termites, this unwanted guest creeps into projects and slowly eats away at the foundations.

Why does Parkinson's Law attack programming projects so heavily? Estimates. Regardless of programming methodology, every project requires some of level of estimation. Without it a project is a rudderless ship heading towards a rocky coast. Estimating is not only a necessity, it's usually the first question asked of a developer: "How long will it take to get this done?" The estimation process is a common place where projects can loose momentum due to Parkinson's Law. It is important to have a clear, defined, and efficient process for estimating tasks. Tasks should be broken down to their proper level of clarity. This level will vary by methodology and team. Enormous tasks and distinct deadlines are a breeding ground for Parkinson's Law. Tasks should be placed into short cycles for completion and review. A consistent cycle helps to maintain focus on the tasks at hand and provides a clear picture for what is next. The Agile and Lean methodologies both focus heavily on these concepts. When defining a task, be sure to provide a concise definition of what needs to be completed. Achieving this may require stating what the task is and is not. This includes defining when a task is considered complete. This can be individualized per task or a larger concept for a project. Think of a task as a 3D box. All sides must have a defined size as well as what is inside and outside of the box.

Creating incentives is an excellent way to combat Parkinson's Law and help a team meet an important deadline. These incentives need not be monetary, but they must be meaningful. Improper execution of incentives can create dissension and dissatisfaction within a team. When defining an incentive be clear on expectations of output, quality, and timeliness. Do not overuse this concept, as it can backfire if a team begins to feel a sense of entitlement.

Finally, another way to combat Parkinson's Law is to find opportunities to positively challenge developers. Ask a developer to estimate a task, then ask if it can be done quicker or if it can be completed in a different way. Challenge a team to take on one additional task per cycle. See if the results were positive or negative. Always look for positive ways to challenge a team or individual. Challenging individuals is about personal growth, not a company's bottom line. This growth will shine through both the developer and team. One word of caution: similar to incentives, be weary about overuse.CodeProject