If You Can't Measure It, You Can't Manage it!
Is your S&OP process working? In this issue of the DBMExecutive I will cover three basic measurements that will answer this question: Bookings Performance, Shipment Performance and Supply Performance. I refer to them as the three S&OP Effectiveness Measurements.
As each month is closed, you get a chance to compare the actual performance during that month to the plan that was created in the previous month. If the actual was within the tolerance bandwidth of the plan it was a good month. If it was outside the bandwidth then it was not a good month.
The chart below is an example of a summary presentation of S&OP effectiveness. If a cell is green, the actual results were within tolerance for that month and a red cell means they were not. For example, the bookings for Family 1 were outside of the tolerance range in January, February, March, June and August. They were within tolerance the other seven months. Family 2 was not reported until March and Family 3 was not reported until April.
This chart gives a summarized visual overview of how well the S&OP process is running. It should be covered at the beginning of every Executive S&OP Meeting.
There is also a story to be found by tracking the trend of a specific measurement within a family. In the following graph, you can see the booking results for a family. The grey area on the left half of the chart is history and the white area to the right is the future plan. The red lines represent the previous month plan plus and minus the tolerance, which in this case, is 20%. The blue line represents the actual results in the grey section and represents the new plan in the white section.
When the blue line goes outside of the red lines, the process is out of tolerance. You can see that the results were out of tolerance in March, June and November (these results will show as red cells on the summary chart). You can also see that the future plan is going out of tolerance in June. Notice that the tolerance bandwidth opens up in the future, as this family has more planning flexibility beyond the first two future months.
This example is also indicating a bias in the future plan. In general, all of the months are higher in the new plan than they were in the old plan. The next graph shows the same presentation with the addition of a trend line for the new plan. Bias can have a cumulative effect on the supply chain but that is a topic for another article.
So, now we have the question of how to establish tolerance levels. Obviously, the wider the tolerance bandwidth, the easier it will be to stay in tolerance. In the field of Statistical Process Control, the upper and lower control limits (the red lines) are determined based on the demonstrated capability of the process. While there is a standard SPC logic to calculate the control limits, I just use a rough rule of thumb for setting the tolerance based on one failure in the last 12 months.
In the previous example, we had a demonstrated performance of three failures in the last 12 months. This would indicate that the tolerances are too tight based on the existing capabilities. If we open the tolerance up to plus or minus 25% there would have only been one failure. The next graph shows this:
Now we only have one failure or out of tolerance condition (June) in the past 12 months. Plus or minus 25% is closer to our real capability of forecasting bookings one month in the future. To address this increase in tolerance you will need to adjust your supply chain strategy. You might decide to increase your buffer inventory or extend your lead-times when you have higher than expected booking months. The tolerances are real and you have to plan for this variability.
To change the tolerances you have to change the process. If you develop better planning processes and demonstrate tighter results, then you will be able to tighten the tolerances. They are what they are until you improve, so plan for variability.
If your S&OP process is mature and you have addressed the tolerance issues, the summary chart should look something like this:
Here is a quick review:
Set initial tolerances. The tolerances for bookings are usually greater than the tolerances for shipments, which in turn are greater than the tolerances for supply or production.
Measure actual bookings, shipments and supply (production) against the tolerances and graph the results.
Display the results in a summary chart with red and green cells like I’ve shown you.
After 12 months, rework the tolerances based on the last 12 months of demonstrated performance. If you improve the process, you should be able to tighten the tolerance but until then, you need to plan for the demonstrated variability.
This will give you a good starting point for measuring if your S&OP process is working. Give it a try and feel free to contact me if you have any questions at email@example.com.
To find Duncan's book, you can go to Amazon.com.