Appropriate inferences and decisions require the best estimates and understanding of sources of variation, necessary for comparability, establishing process control systems, and understanding the process design space. This includes selecting the correct variables, good experimental designs and appropriate analysis. However the concept of components of variation can be confusing to non-statisticians and can lead to non optimal use of resources. The problem is made more difficult when encountering complex and non-linear processes that result in non normal posterior distributions. The challenge is to bridge the understanding between the statistics discipline and subject matter experts, ensure efficient deployment of resources, and ensure appropriate methodology with analysis that is easy to follow and review. What are the challenges and successful strategies in a regulated environment?