Primary Metric(s) – Meaning and its Characteristics

In order to truly control a process, one must understand the inputs as well as outputs of the process. These inputs and outputs are displayed through numbers which are called “metrics”. The primary metric is a way to say with numbers how well your process is performing. Here are a few pointers that will help understand the concept of primary metrics in detail.

What is a Primary Metric ?

Any process can be defined as Y = f(x) where X’s are the critical inputs. The output here is Y. The primary metric of a process is the Y of that process, meaning an output of the process under consideration. The state of the output needs to be consistently measured, errors must be spotted and correction must be immediately made. It is therefore of vital importance for a Six Sigma project to identify the correct primary metrics. This ensures that the project is in control. Choosing the wrong primary metrics will create a situation wherein the wrong thing is being measured. This focuses the organization in the wrong area creating colossal losses in the long run.

Traceable to Critical to Quality Measures

The primary metric does not appear in this stage. It has been present in the project since the first step. The primary metrics have their roots in the critical to quality measures defined by the customer. The get recorded in the problem statement, goal statement and business case. Then they are formally chosen as the primary metric if the project is being correctly executed.

What are the Characteristics of an Appropriate Primary Metric ?

• Accurately describes the desired condition: Metrics describe a condition in the process. For example in an automobile factory, the number of cars being produced per hour is a measure of how efficient the assembly line is. Let’s assume that the number of workers present in the factory was chosen as a metric, it would cloud the judgement of the management. This is because it is possible for all the workers to be present and not produce a single car. Hence it is a bad metric for efficiency. The metric must accurately describe the state of affairs.

• Time lag should be minimum: Metrics need to be seen by the management to take corrective action if the project is going off track. The sooner the management finds out this information the better it is. If the metric takes a considerable time to calculate, it does not solve the purpose very well. By the time the management finds out that the process is out of control, considerable damage may have already been done.

• Not open to manipulation: Supervisors usually have their own agenda. They are likely to manipulate the metrics to get the additional bonus or get shielded from the punishment. If the metric allows people to behave in such in a manner, it has not been efficiently chosen. A good metric is non- human i.e. automatically records and transmits information minimizing corruption.