The Canary in the Coal Mine: Backlog and the Early Warning System
Updated: Aug 27
This post was originally published in DBMExecutive 4.03 on December 8, 2011. The PDF version of the DBMExecutive here.
When we talk about S&OP, we talk about managing the business. We forecast bookings and shipments and put a production plan in place to meet these forecasts. The difference between the shipment plan and the production or delivery plan is inventory. The difference between the booking plan and the shipment plan is backlog. Both of these are strategic “levers” that can be used to manage the business.
Inventory is a lever in S&OP to balance demand and supply. We can plan a level of inventory to meet a desired service level. We can flex the level of inventory up and down to buffer ourselves from fluctuating demand and maintain a level production plan. We can increase inventory in anticipation of plant shut downs. We can measure the results of our inventory reduction plans in S&OP and balance this against our customer service metrics by family. As a lever to run the business, inventory is pretty well understood.
Backlog is also a lever to run the business. The challenge is that this lever is not often as well understood as inventory. The main reason for this is that most S&OP processes start from Operations then move into Sales. Operations understands inventory because it’s tangible and easily measured. When we have too much, our financial measurements show it.
The level of backlog is important because it impacts production efficiency, supply chain inventory, and customer service levels. If your backlog is too high, it will have a negative impact on customer lead-times. Incoming orders must be promised further out in the future because you have too much demand to meet with current production. In this situation, companies typically also experience operational inefficiency as they try to reshuffle orders trying to meet customer demands. As priorities change on the orders, your supply chain struggles to keep up with the shifting demand. Conversely, if the backlog is too low, Operations is looking for orders to build. Lines are run at less than optimal levels and typically are flexed up and down based on incoming orders. The supply chain has to react to fluctuating production levels.
But how much backlog do we need? How much is too much or not enough? The answer to these questions is: It depends. Now before you stop reading further, I will spend the balance of this article answering this question. The answer will still be “it depends”, but you will understand what it depends on, and how to calculate an appropriate backlog level for your business. First, let’s start with some definitions for clarity in understanding this topic.
Total Backlog is all open customer orders that are in your system. This includes both past due and future dated orders. For S&OP purposes, it’s important to see the total backlog as well as the time-phased backlog. Time-phased backlog is a view of the open orders in the month in which they’re due. There are three different types of backlog:
Past Due Backlog: Sometimes we call the overdue customer orders “bad backlog”. These are orders that you should have shipped the previous month but didn’t.
Current Backlog: For the purposes of S&OP these are customer orders that you can ship in the current month. These are all orders dated in the current month PLUS the overdue orders from the previous month(s). All of these orders can be shipped in the current month.
Future Backlog: These are orders that customers want out in future months beyond the current month.
The Current Backlog ratio is the critical one for S&OP. It’s the level of shippable orders in the current month that will impact your service levels and shipping performance. The level of current backlog that you require depends on three factors:
The mix between Make to Stock (MTS), Assemble to Order (ATS) and Build to Order (BTO) in the S&OP family. The mix I expressed as a percentage of the total sales for the family. I usually start by looking at the last 12 months of history to determine the percentages. Remember, you may not have all types of products in a family.
The market lead-times for each product type. Think of these as the external lead-times for the family. How soon after the receipt of an order can the customer expect the items to ship? You should have this documented in S&OP as part of the family characteristics.
The total monthly demand for the product. In this case, demand refers to shipments and not bookings. Once again, you can get an average monthly demand number by looking at the last 12 months of history.
Using these factors, you can calculate a target backlog level. Let’s work through an example. The family we are looking at has average monthly sales of 12,000 units. The family has all three types of products in it: MTS, ATO and MTO. We analyzed historical demand and determined that MTS items accounted for 30% of sales, ATO is 20% and MTO is 50%. From our marketing literature, we were able to get the published lead-times for these types of items: MTS – 1 day, ATO – 7 days, MTO – 35 days. Using this data we can calculate the average monthly demand for each type. I’ve summarized this below:
The next step is to determine the shipment rate per day. To do this, I’ve assumed an average of 22 shipping days per month. The daily shipping rate times the market lead-time will give me the target backlog. I’ve extended the table to show this:
There is a catch. You may have noticed that the MTO items have a lead-time longer than a month. This means that all of this backlog should not be required in the current month. Theoretically, we should have some of the demand for the following month already sold coming into this month. The amount would be the difference between the 22 shipping days and the 35 days of lead-time or 13 days of shipments. By expressing the ratio of shipping days to lead-time for items with lead-times longer than one month, we can do this calculation. Putting it all together, we get the following table:
For this family, our target backlog quantity is 6,928 units or 58% of the monthly planned shipments. An important note on this target: It is better to set it as a ratio or percentage rather than a specific number. The reason for this is that the ratio depends on the mix between the different types of products (MTS, ATO, MTO) and will stay the same as long as the mix is relatively stable. The level of backlog will change depending on the level of demand for the family.
If our S&OP shipment plan is going up to 14,000 units for the next month, we should have 58% of the month already sold. This would equate to a backlog of 8,120 items. This is your target current backlog.
Compare your starting current backlog for the month with this number. If your current backlog is much lower than this, you have a higher risk of not hitting the 14,000 units in shipments. If it’s much higher, you may over ship in the month or extend lead-times. To visually depict this as part of your S&OP presentation set +/- ‘X’ percentage tolerance on the target. If the actual is outside this tolerance, flag the current backlog number as red. Your shipment plan is at risk.
Using this methodology, you can set the target backlog percentage for each family. To start, do the analysis to determine the mix of product types in each family and gather the customer promised lead-times. Once you’ve calculated this, look at how your results match. Does a red measurement result in a missed shipping plan? Once you’ve set the target, you should recalculate it at least quarterly to account for mix changes between MTS, ATO, and MTO items. This is especially effective if you are introducing new products into a family.
Aim to keep your backlog within the target range and you will increase your ability to meet the shipping plan and keep your customer happy. So the answer to how much backlog is still “it depends”, but now you know what it depends on.
To find out more about managing backlog. View the Backlog Video on YouTube.