Welcome to our blog. This is an archived post, most of our knowledge and advice remain valid but some material or links may be outdated. Click here to see our most recent posts.
In this blog I would like to demonstrate a way of using performance indicators to improve process performance, as I did in one of my projects. In a previous blog I explained what a performance indicator entails and presented some practical tips for using them. Now I would like to demonstrate the value of using PI’s in practice.
The project is performed within a large organization in the public utilities sector. This organization consists of several business units which are geographically dispersed, while they are supposed to execute their business processes in a similar way. However, the monthly PI reports showed some major differences in the performance of these business units that were hard to explain. Our challenge was to provide insight in the business processes of the several business units and to declare the differences in the performance of these business processes. In the end, this should contribute to better aligning the performance of these business units.
Our approach consisted of three steps:
First we started modeling the business processes. When capturing the business processes we soon discovered that on a logical level, no major differences in the business processes existed, nor differences existed in the way how the PI was measured. The devil was in the details; the level on which we did find differences between the business units was on the level of working instructions (e.g. booking instructions, administrative procedures, guidelines to be followed). This resulted in the differences as shown in the PI-reports.
To identify the causes for difference in performance, we analyzed PI reports of the last four months. We realized that this was an efficient way to retrieve the main causes responsible for eighty percent of the differences in performance. To do so:
We related the PI’s to the processes described in the first phase and figured out how the data used for the PI’s was generated in the process.
We selected key candidates for interviews (those business units that scored significantly high or low on a particular PI).
Based on our PI-analysis we conducted interviews with the employees from the several business units. In each of the interviews we zoomed in at specific parts of the process, depending on our findings in the PI analysis. For all of the interviews, the processes described in the first phase of the project were used to map possible causes for differences.
After the interviews, we organized sessions with the process analysts of the organization. In these sessions we analyzed the outcomes of the interviews. From these sessions we could draw two main conclusions:
At the level of work instructions and guidelines, some major differences exist in how things were done in the several business units. For example, the way how people register orders in the system (booking instructions) and the way how business units prioritized the work to be done (which resulted in differences in cycle time of particular orders). These findings were mapped on the process model and we were able to come up with an initial draft of a ‘best practice’ process (including guidelines, basis for work instructions, policies, etc.).
At the same time, we found some causes to declare the variations in the PI reports that were not generated by the way the process was performed. For example the geographical position of a business unit (and hence the access roads in a particular area) resulted in differences in the lead time of recovering breakdowns in the infrastructure (which was a PI). These results were presented to the management in order to get the necessary insight for interpreting the PI reports.
PI’s are a useful means to get insight in the performance of processes. Even when we have a clear understanding of how the process is designed, the actual execution of the process – and hence its performance – can differ from the way it was intended. Partly because people can deviate (conscious or unconscious) from the way how the process was designed, but also contextual factors (e.g. geographical location) can have a major influence on the performance indicators of a process. In the above described project, PI’s proved to be a useful means to figure out the differences in how processes are executed within multiple business units. An important remark should be made: be cautious with blindly relating PI scores to how a process is designed. Deviation in process performance – as shown by PI’s – can be caused by other factors than design of the process itself (such as human defaults, contextual factors and available resources). Solely relying on PI scores, may lead you to wrong conclusions. PI reports are a useful means for analyzing the performance of processes, but to really improve this performance, the prerequisite is the conversation that follows from these numbers!
SUBSCRIBE TO BIZZDESIGN'S BLOG
Join 10.000+ others! Get BiZZdesign's latest articles straight to your inbox. Enter your email address below: