When we are talking about data driven business or in general about industrial processes, we are talking about numbers and performance. To measure the performance of processes and keep control on them we need appropriate metrics, perhaps a complete set of metrics, a metrics system. These metrics, if carefully chosen, provide deep insights in your business and will help you to identify trends or to identify deviations from your normal (or planned) performance. But which metric do you have to choose? What are good metrics for your business?
In general there are four attributes a metric should have to create a good and meaningful metrics system for your business:
- A metric should always be a rate or ratio (e.g. scrap rate or First Pass Yield in production). Never ever pick an absolute number for a metric. You will never be able to compare it over time or compare similar processes.
- A metric always should be connected or related to other metrics. In ideal case you can calculate your metrics from each other by simple arithmetic operations like + – x /. If you really know the relationship between your metrics (and therefore you have understood your business), you should be able to arrange your metrics in a metric tree (see also picture below).
- A metric should give you control over your process and targets for this process or at least a part of them. In other words, the chosen metric should influence your behavior if something is changing in a not desired way or moves in the wrong direction.
- Twenty is plenty! Do not choose too many metrics. Too many metrics will make your business diffuse and it is not easy to decide which of the metrics have the highest priority and where to set the most impacting action.
Which metric you choose to handle your business in detail highly depends on the business you are working in and which processes you want to control. But if you follow these basic rules described above you are on the right way. Start with your highest level of “Key Performance Indicator” (KPI) and try to break down in sub metrics together with your team. Always keep in mind, that your metrics (respectively their targets) will influence other processes along your value stream. For example, if your production metrics show low performance, this might influence also the sales department. Always try to align local metric systems with your internal and external business partners.