“85% of the reasons for refusing to fulfill customer expectations you can trace back to failures of systems and processes, and less to people. The task of management is to change processes, not the people.” (W.E. Deming)
Mr. Deming’s opinion is that it is your processes and not your employees that possibly spoil your profit. Is this a legitimate view or rather a somewhat closely seized opinion? When William Deming – the godfather of the Japanese Wirtschaftswunder – was in business, processes mainly existed in the context of the producing industry. Assembly line enterprises were (and still are today) run in the spirit of the “scientific management “a’ la Frederick Taylor (who by the way was fired for his mismanagement at Bethlehem Steel). Note that the assembly line metaphor is a powerful marketing instrument. It suggests that business processes can be as easily controlled as an assembly-line with electric motors, quality checkpoints, checklists, and its specialized and exchangeable workers. THAT is what every process expert dreams of.
However, the reality of a software project looks different. The magnitude of possible choices during the product design is of immense complexity. A certain process definition may serve well in the current project. In a different project, the same process definition might have catastrophic consequences. As an attempt to deal with this problem, the idea of so-called “Tailoring Guidelines “ (as defined in the CMMI model on the third maturity stage, for example) was created. The “Tailoring Guidelines” are a part of a meta-process, which purpose is to allow for a more flexible configuration of the basic production process definition. As a result, the first derivative of the failure was created here. In everyday practice, this construction appears quite abstract and rather difficult to understand. It takes a long time (months or even years) to prove its effectiveness. In the meantime, the responsible process advisors have long left the scene and their names have already decayed in oblivion.
It is not the first time in the history of Information Technology that doubts arise about the practical use of theoretically convincing models. In Mind over Machine, Dreyfus notes that it seems impossible to create a computer model of the human decision-making process. Just think of the artificial intelligence hype of the late 80s, were the “computers of the 5th generation“ were supposed to become intelligent machines. Are we making the same mistake with the formal process definition as was made with neural nets, fuzzy logic, cybernetic machines, and genetic algorithms? Are we trying to capture the incomprehensible? Perhaps we got lost again in a hopeless attempt to reduce everything to just one single aspect? Like focusing on abstract “processes” instead of dealing with real human beings who actually constitute these processes?
“The task of management is to change processes, not the people,“ argues Deming. Maybe he said that because he recognized that it is much easier to define a multidimensional process model for a whole automobile factory than to change the ways of just one human individual?
It’s not to say that it is useless to (reasonably) define business processes. We simply don’t know a better way to effectively control a complex enterprise. However, we should be very careful and proceed pragmatically. Experience, extensive knowledge, goodwill, and goal orientation are needed to define and improve a business process. No theory supplies a safe prescription for a successful business process optimization. Quality assurance models like CMMI can merely be viewed as a help for us to stay on track. Only a professional consultant with the right know-how will actually be able to optimize your processes. In this regard, let’s just hope you have found the “right” advisor for your project.