In the late 90’s, the Six Sigma quality improvement process swept through the corporate world, led by early and vocal adapters such as Motorola and GE. For those who are not familiar with Six Sigma, it is a process for making significant quality improvements through a rigorous five-step data- and statistical-based approach (Define, Measure, Analyze, Improve, and Control). The siblings of Six Sigma, such as Design for Six Sigma (DFSS) and Lean Six Sigma, expand its applicability to new product design and manufacturing flow. You can immerse your products from cradle to grave in the Six Sigma process to achieve total quality control nirvana.
While this may be great if you have the resources of a GE, and are involved with markets and products that are more mature and slow to change, I’m here to tell you that Six Sigma is anathema to the nature of the startup. Of course, I’m not proposing that quality is unimportant. What I mean is that the rigor and thoroughness of a true Six Sigma process goes beyond quality control into a near-obsession with perfection. Such an approach, which some startups do adapt, can lead to “death by measurement” or “analysis paralysis,” where more time is spent measuring results than actually achieving them. This works against most of the attitudes needed to make a startup successful. For some specific examples:
- Failure – The goal of Six Sigma is to avoid failure or a product or even a single aspect of a product at all costs. The price you pay is time. This runs against the startup mantra of “fail early and often” – in other words, the belief that you don’t know what will be successful until you try, so you better try a lot.
- Precision – Six Sigma places a lot of importance on having a quantifiable understanding of customer needs (translated into measurable technical specifications) and sufficient data to statistically prove (or, at least, being unable to statistically disprove) conformance to these specifications. Gaining sufficient data requires a large sample size which again requires time and, in this case, cost that you can’t afford.
- Perfection – The Six Sigma process is aimed at delivering six-sigma quality (3.4 defects per million). Unless you are landing airplanes or performing brain surgery, you probably don’t need that. Better to move onto the next generation / iteration than to spend time perfecting the existing one. By the time you do, the needs of your customers will be better filled by someone else who didn’t spend the time.
And another, perhaps even more important point (even for larger and more mature companies) is covered in this Industry Week article warning that Lean and Six Sigma can cause an “identity crisis”. The view of the author is that these processes are very internally focused and not well-suited for building and maintaining close customer relationships. As a certified Six Sigma Green Belt at one of the major corporate believers in Six Sigma, I can vouch that many times the pressure was just as much on broadening the specifications to improve our process capabilities as it was on improving the process itself. Some of the same points regarding speed and adaptability to changing conditions are also raised in the article.
There are obviously several good elements of Six Sigma that do apply for startups – for one, implementing controls (which was often the downfall of even well-executed Six Sigma quality improvement efforts) is important for creating consistency in any product or process. Additionally, the attitude that critical to quality parameters (CTQ’s) can and should be measured is also helpful to maintain (and that the measurement systems must be capable). But beyond that, you are better served in a startup by focusing your “quality” efforts around really understanding what your customers perceive as quality, and how it is changing as they are exposed to new ideas.
Image credit: Brooks Elliott