Here's the juice - a condensed version of Six Sigma training principles. There are five critical stages - Define, Measure, Analyze, Improve, Control.
1. DEFINE the problem
- What has gone wrong?
- What is not right yet?
- What is the customer not happy with?
- What are they complaining about?
- Where are we falling short?
- What are the things we want to change?
- What are the things we need to change to survive?
Identify the “CTQ” – Critical to Quality factor – that your customer will really value.
The key at this stage is not to “boil the ocean”, to think that you can fix everything all at once. Break down the problem – chunk it down – into bite-sized and manageable pieces. Then focus all your resources on achieving that one result, because focus is power.
2. MEASURE the problem
- Does the measurement system work?
- How is the process performing?
- Draw a process map so you can visually see where the
- Where are the bottlenecks occurring?
- Consider using a formal Failure Modes and Effects Analysis (FMEA) to assess the severity, occurrence and detection rate of every problem.
Work out your Defects per Million Opportunities (DPMO score) and then convert to a Sigma score (there are plenty of books and sites on Six Sigma which will show you how to do both these calculations).
3. ANALYSE the problem
- The famous fishbone diagram is very useful here. List the result (the "effect") at the head of the fish and then all the root causes as the bones feeding into the body of the fish. Try a simple cause and effect matrix, giving numerical values to the importance of every input in driving the output.
- If you want to get technical, use hypothesis testing to rigorously analyse the data – “torture the data until it squeaks”, we used to call it. The myth is that analysis requires a PhD in statistics. In fact, 80% of all problems can be solved by a straight forward review of data and applied common sense.
- Look for basic correlations - but remember that correlation doesn’t equal causation. Consider drawing a basic Pareto Chart so you can see just how 20% of causes result in 80% of the problems. What is normal common cause (random variation) in a process, and what is a special cause (something that drives major variation?)
4. IMPROVE the situation
- Brainstorm and decimate the problem. Imagine every problem has a hundred solutions and act “as if”. Funnel solutions using cost-benefit criteria and think practicality.
- Go for simple solutions that can be easily implemented. Use pilots and designed experiments to test what works and what doesn’t. The ideal are failsafe, poka-yoke solutions e.g. bar codes on products, back-up power on alarm clocks.
- Measure afterwards to verify the solution has actually worked – never, ever assume.
5. CONTROL the process
- People ask what the hardest part of Six Sigma is, and I reply that it’s keeping the process under control.
- Use control charts to track performance. Implement a Control Plan and a Reaction plan.
- Make the process owners accountable for the process performance, and emotionally bought-in to its continued success. Six Sigma you see, is not about rocket science. It’s about people.
That’s Six Sigma Training in a nutshell. I hope this has stripped away some of the mystique and smoke-haze around the concept. Six Sigma is about quality, and the pursuit of excellence, and unlike business fads and jargon, that will never go out of style.
(c) WestOcean 2009
