Business process management is an approach that seeks to understand, improve, control, and automate business processes to improve process performance. A crucial part of business process management (BPM) is Process Behavior Charts (PBCs), also known as Shewhart charts or control charts.
Process behavior charts are used in quality management for production, customer service, and administrative processes to monitor critical process performance indicators. These charts focus not on individual product defects but on the behavior of the process itself over time.
Here are mistakes to avoid when using PBCs.
Failing To Understand The Nature Of Variation In A Process Before Collecting Data
According to reports, the business process management market was valued at $3.38 billion in 2020.
Process behavior charts are only effective when in-control or “common cause” variation is measured. For example, if the number of defects is inconsistently high due to disturbance factors such as variations in the materials, the operator’s mood, or other random factors, PBCs would be of no use.
Failing To Collect Data For A Sufficient period Before Drawing Conclusions
The control limits on process behavior charts are based upon three times the common cause variation in a process.
If you want to have 95 percent confidence that your control limits will not be exceeded, you must base the limits upon at least 27 data points acquired over some time long enough to include three complete waves of common cause variation.
Setting A Control Limit Too Wide Or Too Narrow For Your Process
The control limit lines on a process behavior chart are one standard error away from the centerline, the process average.
If your control limit is one standard error away from the centerline, you will have 95 percent confidence that your control limits will not be exceeded. If your process location (average) shifts due to special causes of variation, this shifting will be reflected by the position of the control limit lines.
Failing To Monitor Subgroup Size
If subgroups are too large, you will not be monitoring the performance of individual parts or people. If subgroups are too small, you will not have enough data points to establish control limits (typically three times the common cause variation) and detect special causes of variation.
Misinterpreting Control Chart Signals
A process behavior chart will indicate when a process is under control, not necessarily performing well. A process may be in control, but its performance could be poor.
A process without special causes of variation is in control and provides information about the collected data’s average or central tendency. However, this type of chart does not indicate the dispersion of the data.
Taking Action To Improve A Process Without Establishing If It Is Under Control Or Not
Without establishing whether a process is under control, you have no way of knowing if any improvement being implemented will result in better performance. If a function is not under control, you need to correct the situation, or advances will result in erroneous outcomes.
Getting Too Concerned With Small Shifts In The Process Location (Average) And Ignoring Large Shifts
When a process is under control, you expect to see variation between data points due to common cause factors such as natural variation or changes in materials or equipment.
If your control limits are set correctly, this type of variation will result in some average shift over time. If your process is not under control, it is doubtful that you will see an average change due to common causes.
Creating A Chart But Never Using It
If you collect data using one process behavior chart and then start using another tool to monitor the same performance measure, the first chart may not be helpful anymore since it is already outdated.
Failing To Use Upper And Lower Control Limits
A process behavior chart has three lines on it, not just one. Some process improvement models suggest only using one average line on a process behavior chart. The practice can lead to false alarms of potential problems in your process when there are no problems present.
Overreacting To A Special Cause Of Variation And Failing To Investigate The Reason For The Problem
If a process behavior chart indicates a particular cause is present, you need to take action or risk having this situation recur immediately. If the system is out of control, it is better to be safe and take pre-emptive action rather than wait until it exhibits a problem once again.
Assuming Your Process Behavior Charts Are Set Up Correctly The First Time
Some people believe that setting them up will become second nature when using control charts for a long time.
It is not necessarily true. Setting up control charts may take some time to get it right, but they should be put away after this is done.