As an analyst details matter. Accurately reflecting reality through numbers, charts and visuals becomes a passion. As an engineer details matter. Designing a system or method to improve the manufacturing or reduce the variation of a product becomes the goal. Details are vital to both endeavors. When dealing extensively with details, the strong nagging question of that elusive detail or overlooked variable is always present. This can easily lead to analysis paralysis. While getting the details correct, the goal of gathering those details must include an action.
Luckily, a couple different models provide a method to break through this impasse. The models are designed for major projects, but the concepts work for any moderately complex work. From computer system analysis, the spiral model approaches system development in a series of structured developments. Initial planning of the project including some key milestones precedes an analysis and design phase. Determine the area that presents the most risk and address those concerns early in the project. Once those risks are understood begin designing a system to address the most important concern. As the design finishes, implementation is ready to begin. Implementation starts with a working prototype to make sure each function works properly and then testing and integrating the new system. Once that is complete, start the process over again for the second iteration. The focus should be risk mitigation and avoiding analysis paralysis.
In quality, the Deming Cycle of Plan-Do-Check-Act provides a method to systematically improve a process. Like the spiral model, Plan-Do-Check-Act works through a series of steps to continuously improve and avoid analysis paralysis. Unlike the spiral model, the focus is to improve a process. The Planning activity must be designed to lead into actions. During the analysis, it is critical to understand the perfect knowledge of a subject cannot be attained. Obtain the necessary information to decide. Make a pilot of the improvement where most of the variables can be relatively controlled. The pilot should contain all the key functions identified in the planning phase. Check the progress of that pilot, make any necessary adjustments if the project essentially meeting the objectives. Finally act on the checks that were performed. Hopefully the action is to replicate the project to a wider area or across the company, but occasionally the action is to discontinue the project. Either way, the information found in the pilot is not wasted. Work that information into future projects.
The spiral model and Deming Cycle are two methods to avoid getting mired in analysis. Another important concept to remember while performing the analysis is distinguishing between maximizing utility and satisficing. Maximizing utility is getting the very best out of every decision while satisficing is attaining a good decision even if a marginally better decision could be made. Finding the best solution sounds like the ideal approach to either method, but as the analysis is prolonged searching for the maximum utility. Satisficing gathers the information necessary to make a good decision and proceeding to the next phase of the project. When looking at either cycle, the goal is moving to the next step with the necessary objectives met.
For those analysts and engineers out there that struggle in the analysis, consider the objective of your analysis. The analysis is not an end to itself but a necessary step to planning. When sufficient analysis has been done, make the decision with conviction.