Lean Six Sigma: Are DMAIC Demons Possessing Lean?
Halloween is a few weeks away, and the strange world of the dead, undead, partially dead, or dead fashion is inhabiting the shopping malls. Over the last two years I’ve watched another kind of possession has happened in business thinking. Once the darling of Motorola, GE, Bank of America and others Six Sigma has been falling out of favor. Six Sigma’s DMAIC model, its emphasis on gathering statistical data and focus on elimination of variation has floundered in knowledge work. Knowledge work has at its heart, people. There is no machine stamping parts. In knowledge work, the work being done is continuously varying in size and complexity. The work of problem solving is done in people’s brains, and the effective flow of information in the work system controls errors. All of the variation in these complex adaptive systems means trying to eliminate variation is a zombie’s errand. Lean and the Toyota Production System are principles based. Many of the ideas in Lean can be applied to Knowledge Work. Ideas like Respect for people, Pull, Flow, Visual Management, Value Stream Mapping, Systems Thinking and Continuous improvement all make sense in Knowledge Work. These Lean and TPS ideas can be used with systems that adapt and change based on the variation inherent in the work.
Lean Six Sigma
Over the last few years Six Sigma consultants have started to re-brand themselves as Lean Six Sigma consultants. The adoption of Lean onto their business cards is partly a survival strategy. If Six Sigma’s not selling and Lean is, then let’s “do” Lean too. The problem is that Lean and TPS come from a very different perspective on work than Six Sigma. For example Six Sigma’s DMAIC process is:
- Define the problem, the voice of the customer, and the project goals, specifically.
- Measure key aspects of the current process and collect relevant data.
- Analyze the data to investigate and verify cause-and-effect relationships. Determine what the relationships are, and attempt to ensure that all factors have been considered.
- Improve or optimize the current process based upon data analysis using techniques such as design of experiments, poka yoke or mistake proofing, and standard work to create a new, future state process. Set up pilot runs to establish process capability.
- Control the future state process to ensure that any deviations from target are corrected before they result in defects.
There are some major problems with the DMAIC approach when dealing with knowledge work:
- Defining the problem with voice of the customer is great. However what if the customer can’t articulate what they want? Every software customer I have worked with did not have a functional crystal ball. They can often describe at a high level their needs, but they do not yet have enough information. The only certainty is that they won’t have enough information. Neither do developers.
- The underlying assumption of Six Sigma is that the process under study is a repeatable process producing the same type of work product and subject to some statistically relevant variance. Measuring aspects of the process gives data to better understand where variation occurs. In Knowledge Work such as software development or marketing campaign creation the work is continuously varying so the measurements measure a complex adaptive system, not a repeatable process.
- Analyzing the data gathered will provide some insights. Let’s say that data reveals a problem and a solution is found. Putting an improvement in place will definitely help. We may rarely do that type of work again. What happens when down the line a different problem occurs that we have never seen and do not have a measurement for?
- How do we know we have the most correct process just from analyzing work in our current process? In Knowledge work, the work itself often impacts the process.
Six Sigma’s underlying assumptions fail to account for complex adaptive work systems that rely on people to do the problem solving and rely on information flow to do error correction. So why do people apply Six Sigma processes to knowledge work and what happens when they do? More on this later…. time to catch a plane.