The use of Statistical Process Control, or SPC, is a technique that examines a set of measurement values from a process to determine the arithmetic mean of the values, and establishes upper and lower control limits for the process values. Control limits basically show the upper and lower bounds for the values or results from a process. These limits represent the bounds of normal variability based on a set of values used and input to the SPC exercise. Values above the upper control limit or below the lower control limit are labeled as outliers. Without getting too deeply into the statistical mathematics, the control limits represent 3 sigma from the mean, and according to the statistics, 99.7% of all values should fall within the control limits. Often, additional limits are identified that are 2 sigma from the mean, and these are labeled warning limits.
The figure below, what is called an SPC Chart, shows the mean and lines for 1, 2 and 3 sigma.
An important point to note is that these lines are dependent on the input values, and should be a representative set of values, as large as possible. Statistical analysis is not going to work over a 1 week set of 5 daily measurements. A better example is the set of Average Speed of Answer (ASA) values by 30-minute intervals over a one month period.
An important task of the leadership team is managing this variability. The tighter the control limits, or the closer they are to the mean, the lower the level of variability. The focus needs to be on moving these control limits closer to the mean and aggressively managing outliers. Another important focus is on improving the mean line over time.
Trends are important in statistics, and in managing a support center. In fact, typically a certain number of points or values that trend in a certain direction cause a redrawing of the SPC Chart. The mean is recalculated as are the control limits. In practice, the SPC Chart can be redrawn at any time, but this is typically done when a set number of values have trended in the same direction, or when a specific project has been completed or action has been taken to improve the operation with respect to the measurement under consideration.
The following SPC Chart shows values over time for a process, shows the warning limits shaded in red and shows one outlier point or value. The chart also shows a point in time in which a recalculation was done to coincide with a change, hopefully an improvement. The mean shifts , once again, hopefully closer to this goal. The shift would be even more positive if the control limits were closer to the mean as a result of the change.
SPC gives management the tool necessary to identify those processes that are performing outside of expectations so that action can be taken, or to identify specialists that are performing outside of an expected range, allowing for coaching opportunities. Outliers need immediate action, while mean line improvements and control limit changes need disciplined project-based actions.
An example will help clarify the SPC concepts and actions that should be considered by the leadership team when analyzing an SPC Chart. The SPC Chart below shows daily First Call Resolution, or FCR, measured for a one month period. The line labeled FCR represents the daily values for the measurement. You will also see the mean line, and the upper and lower control limit lines. There are three distinct periods visible on the chart, showing shifts in the mean and control limits.
There are three important points to understand when viewing this SPC Chart. First, the mean line (FCR) is increasing in value during the shifts. As a higher FCR value is better, this indicates that there is improvement being made in the FCR over time. If the shifts represent changes in the operation introduced by the leadership team, then the changes are having a positive effect.
Secondly, the control lines are moving closer together over time. This indicates variability is being managed out and the customer is receiving a more consistent treatment.
Thirdly, outliers become visible to the leadership team. Questions can be raised regarding the FCR values that fell below or went above the control limit lines. What might have happened during these days. If the FCR was below the control limit, was a new product or service feature introduced without the proper level of training? Were the specialists with the most experience off of the phones for some reason? What if the FCR was above the upper limit? Is this a problem? Not necessarily, but the values do show a greater degree of variability, and those days should be examined to understand why.
Overall, SPC allows management to measure and track the variability in the support center operation and hence the customer experience. Outliers and variability have a direct and significant impact on the satisfaction of the support center customer, and hence all Key Performance Indicators should be analyzed using the SPC Chart approach.
SPC also enables the leadership team to have a better understanding of where to set goals and stretch goals, or at least where not to set them, as discussed in the introduction. If your FCR data for the past year show significant variability, your operation is not well controlled. Variability equates to risk, and when high, you runt he risk of your number for the year coming in somewhere within a wide range of possibilities. Goal-setting under significant variability should address the variability, not the mean value. Once the variability has been minimized, you stand a much better chance of choosing a meaningful goal.