The principle of measurement is one of the fundamental tenets of Six Sigma. You may boast lots of knowledge about your processes, but until you translate that knowledge into numbers through measurement, you’re bound to the world of gut-feel, guessing, and marginal improvement.
Likewise, you may work very hard to bring significant resources to a performance problem or improvement goal, but without measuring your Ys and Xs, your ability to improve will be weak. (This idea is nothing new; British scientist Lord Kelvin brought it up in 1891.)
Taking measurements is a matter of using information and data to quantify the relationship between inputs, outputs, and error in a given system, process, or operating model.
Measuring many of your inputs and outputs may seem impossible, but resist the urge to rationalize yourself into believing that some things just aren’t measurable. In the long term, achieving your goals without the measurement data that can help you is going to be much more difficult.
Mind your Ys and Xs in Six Sigma
Measurement begins with the output Ys and then extends to the input Xs to understand the causes. For example, you’d probably love to have $1,000 in your bank account (the output Y). To measure how you’re doing on this objective, you can check your account each day and see how much money is there. You’ll probably discover that your performance toward this objective is well below what you want.
As you analyze the situation, you’ll discover that the amount of money in your account is a function of how much money you earn, how much is taken out of your earnings in taxes, and how much you spend on necessities. These factors are the Xs.
You can’t affect the amount of money in your account directly; you have to do something to the identified Xs to make Y change. To affect a change in the output Y, you start to measure and control the performance of the causal Xs (perhaps earn more and spend less).
Many people never get past the Y. They watch it hoping that it will change simply by being measured. Such Y-dominated thinking and measurement is an easy pitfall. Consider the company that continues to work harder and force productivity in an attempt to improve results (the Y) without quantitatively investigating the contributing factors to success. This approach has a tendency to self destruct in its blind push toward a goal.
Measures equal data in Six Sigma
Understanding how to measure certain input and output variables is easy because they’re, by their very nature, accommodating of such measurement. Examples include the time it takes to complete a certain job, the number of days that transpire between a customer order and the delivery of the product, and so on. These measures are all numerically quantifiable.
Such measurement-friendly events, processes, variables, and transactions are well supported by certain measurement tools, such as software, a spreadsheet, a clock, and so on. Using various quantitative scales such as width, length, time, rigidity, and density, you can quantify the behavior of many Ys and Xs, and their relationships.
Other Ys and Xs, however, aren’t so easy to measure because they’re not as easily quantified or because the time, cost, and effort involved in doing so is extreme. How, for example, do you measure customer satisfaction or a customer’s opinions?
Even when numbers or direct measurements aren’t available, you can create them indirectly. In this sense, you take a deterministic approach to measurement: You don’t give in to the lack of data; you find the data you need or create estimates in accordance with sound practice.
In these cases, measurement instruments have to be specifically designed, such as a survey question that ranks responses. With such instruments, you can make otherwise qualitative data much more quantitative in nature.
