Key Performance indicators (KPI) and other related indicators are often hard to compare case by case or year on year. One reason is that each process has various aspects from a causal standpoint:
A machinery defect or non-conformance can have many causes. Which one is to be selected as a KPI?
Perhaps the immediate effect of a defect or non-conformance is not so varied but this can vary too
So risk or potential effects for the same defect or non-conformance can vary greatly.
In fact, we spend days or even weeks each year preparing KPI, but do they help the management and decision making going forward ?
Simpler data to reconcile such as data descriptions of spare parts or crew members’ experience descriptions involve less options. Even so we run into description variance requiring an important decision regarding which variances could constitute the KPI.
Integrating varied data points
We often hear about data driven decision making. How can we do this without reconciling data?
We as humans manage to reconcile varied data points because we understand how they are generated. But this is not always consistent in systems as there are often many considerations before we can arrive at a merging of data points.
Integration and Data Unification Are the Same Challenge
Reconciling data and integration have the same set of concerns. It is about knowing how the data generated. So before it attracts attention as a data point useful for decision making, it is important to recognise how it is generated.
Caring about data
For KPI to work, someone has to care. Someone who is committed to the domain and to extracting the data to assist decisions.