Measuring performance is central to the practice of Lean Six Sigma. Within the DMAIC cycle of define, measure, analyze, improve and control, the measure element is very important. We need to determine the specific kinds of company metrics we will use. For example, if we are in manufacturing, we would be interested in scrap rates or cost of poor quality. In a supply chain, we might be looking at on time delivery and order fulfillment. In a financial services organization or a customer service organization, we might be thinking about call handle times or the number of calls dropped. In the medical industry, you might be looking at the average days of recuperation from a given surgical procedure. There are many different kinds of metrics in every industry and performance is important. No matter if your organization is for profit, non-profit or government. The concept of Six Sigma quality relies heavily on metrics and comparing a performance against a mean or target value in a process. At the level of Six Sigma quality, we would only have 3.4 defects per million opportunities for a given product or service. In order to determine if we are performing at that standard, we need a couple of things. First, we need to understand what’s called an upper and a lower specification limit.
They define the range of what’s acceptable to meet customer requirements. This could be any attribute of the deliverable, such as a dimension, an amount of time, the concentration of an active ingredient, or any number of things that are important to the customer. If the upper and lower control limits are set at three sigma, we’re still going to have many defects, especially if we have a high volume of product. That is why many organizations have adopted the Six Sigma methodology. Metrics in Six Sigma are used extensively and they’re used to measure different things about our process. For example, let’s discuss DPU or defects per unit. In the auto industry, one of the things that is tracked would be the number of complaints per order or per car shipped as a way of measuring quality. Defects per million opportunities examines the number of defects across the process and gives us an idea of whether we are within our specification limits, and meeting customer requirements. Another important metric is called First Time Yield or FTY. And here, we might be examining, for example, the number of calls we could handle on the first try at a contact center without the customer having to call back or follow up, because we didn’t do it right the first time. Let’s move on to the next metric called rolled through yield.
In the example of an automotive assembly plant, we would keep track of the percentage of good quality at each station or step in the assembly line, then we would multiply those to get to our value. So if we had a 97% good quality at step one, 95% good quality at step two and 90% good at step three. We’d have multiplied those three factors and we would find out our roll through quality is actually only about 83%. Cycle time is another important metric, because we want to measure the time to perform a given task and how much variation we are seeing in the process. For example, we measure the time it takes from a customer placing a delivery order until they actually receive the product and whether we can get a consistent result from the process each time. Cost of poor quality includes scrap, rework, returns and expediting that can happen as a result of not doing it correctly the first time. All of these metrics are important and you will decide with your team and stakeholders which metrics will be applied to your process, and how they will be used.