Archive for June, 2010

Six sigma in the moose muck.

June 1, 2010

For the past week I have been working in Berlin, NH. Berlin is a city in northern NH on the Androscoggin river and a mill town. Their slogan is “The city that trees built”. The mill is now closed but the culture lives on. Where I eat breakfast, a restaurant in the old train depot (the tracks are gone as well) has a sign out front that states “Nous Parlons Francais”.

Years ago French-Canadians came to work the mills of New England, now with the mill closed the wood is cut and much of it shipped to mills in Quebec. This is not a story unique to Berlin, it is happening throughout northern New England.

We have a place in Phillips, Maine and some nights I have been going there and other nights I stayed at a friend’s house in Berlin. The drive from Phillips to Berlin is two (2) hours, through mountains, timberland, lakes and rivers. To describe route 16 between Errol, NH and Rangeley, ME as isolated would be an understatement.

In making this drive, there has yet to be a day I did not see a moose. As anyone who has seen one will attest, a moose is a large animal, often weighing 1,000 lbs. There have been cases where a car has hit a moose and the driver was killed, the car totaled and the moose walked away.

This morning (Tuesday, June 1, 2010) I left Phillips at 5 AM and arrived in Berlin at 7AM. Along the way I saw six (6) deer and six (6) moose, one of whom was dead in the road in my lane. Fortunately I saw him in plenty of time and was able to avoid hitting him. To give you an idea of size, he was on his side taking up the whole lane. Lying that way, he was taking up more space than me small 2 door Hyundai.

Seeing all these animals got me thinking about quality. Now obviously the moose are doing quite well here, thank you, unlike the fish in the gulf next to BP’s oil well these days. Now I am a big fan of analytics, performance measurement and statistics in general. Now while the wildlife biologists use lots of statistics to measure and track the moose, I have never met a moose who could so much as count, much less compute a Kendall-Tau statistic.

Yet somehow, without the aide of mathematics, the moose has evolved in perfect harmony with his environment and is flourishing.

Which brings us to a key point; while statistics are appropriate for process improvement, they are no substitute for initial greatness. Allow me to demonstrate what I mean by this. Suppose you are an aspiring composer. If your goal is to create short ring tones for cellphones to sell on the web, an analytics program will definitely help you maximize your profit. If on the other hand you wish to compose music to rival Beethoven and Mozart, no amount of statistical process control is likely to be the key ingredient that separates the truly great from the simply good.

Which brings us to the key point. To be successful a firm should start with great people and then measure performance. Measurement is ex poste while greatness is ex ante.