In my October 2016 column Human Data I showed the DIADEM model of data, which not only clarifies how data differs from the uses to which it is put, but also explains how to go about improving the way it is collected and used.
A particular use of data is in corporate indicators—Key Performance Indicators, Business Performance Indicators, and other metrics that form an agreed basis for tracking the operations of an organization. Like processes, these indicators should not remain static, but rather respond to operational challenges that change constantly. So, improvement of data is often about improvement of indicators—and this is a human process, or rather, a connected set of multi-level processes.
This can be understood by breaking down the improvement of indicators into strategic, tactical, and operational layers. At each layer, the data gathered should support specific questions:
- Strategic
- Concerned with high level organizational challenges
- Process users are senior management
- Questions include:
- Which challenges remain valid?
- For which challenges has the diagnosis of where we are as an organization changed?
- For such changes, what are the drivers?
- Tactical
- Concerned with implementation of the policies that respond to challenges
- Process users are middle management
- Questions include:
- To what extent has each policy been implemented?
- To what extent is each policy addressing the challenges it was designed for?
- Do we need to change any policies, and if so, how?
- Operational
- Concerned with operation of policies
- Process users are operational leads
- Questions include:
- Are daily actions coherent with implemented policies?
- Are these actions being taken in an efficient way?
- To what extent do these actions improve operations of the organization?
All too often, organizations make false economies with data—they collect certain data sets, and define corresponding indicators, because “it's always been done that way” and/or because it apparently is the easiest way to satisfy reporting responsibilities. However, data and business processes should form be interwoven to enable continual improvement of the organization. This can be done by breaking down indicators and their uses into layers, then designing human processes at each layer that ask the right questions, and demand the data necessary to answer them.
Speak Your Mind