• Duncan McLeod

Intelligent Demand and S&OP

View this article as a PDF here.


"This article was originally published Nov. 1 2013."

President's Note

In this article, I will address some of the issues related to managing intelligent demand in S&OP. Actually, the first issue is defining what intelligent demand is. If you are supplying components to support your customer's manufacturing schedule, this article applies to you. Classic examples are companies supplying automotive assembly plants, white goods assembly plants or any other customer that provides a schedule instead of specific purchase orders.

Managing intelligent demand in S&OP requires some special considerations. What are bookings and backlog in an environment where there are no customer orders and only schedules? In this article, I write about an approach to managing this type of demand in the standard S&OP process. In most cases, S&OP families will have multiple demand streams, each with different characteristics. We need to be able to combine the intelligent demand streams in order to get to "The Story" for the family.

What is intelligent demand?

We have coined the term "intelligent demand" to represent the case where the customer provides us with a rolling schedule rather than individual purchase orders.

Why call it intelligent demand? Instead of forecasting or guessing at the demand, as a supplier we have intelligence about our customers' requirements. Their vendor schedule is directly linked to our shipping plan and, theoretically, we don't need to forecast demand.

The automotive industry pioneered this approach back in the 1980s.They

built EDI linkages between the assembly plants and their tier one suppliers. The plant's vendor schedule became the tier one supplier's demand plan. This approach was a game changer for the industry and has since become a common approach to building the link between two manufacturing plants where one supplies the other.

The customer does not place orders in the intelligent demand environment - they transmit schedules. These schedules will be updated and retransmitted daily or sometimes weekly. Unfortunately, the retransmitted schedule may be significantly different than the previous schedule, leaving the planner at the supplying site to question the validity of the term "intelligent demand".

Intelligent demand is dependent demand - demand that is directly related to the demand for parent products (check the APICS dictionary for a full definition of dependent demand). This case is slightly different, as the demand is not driven through the bill of material but through a plant-to-plant schedule link. In following diagram, the finished product, "RY1OO", uses a Wheel Bogey, 2000. The RY1OO is made at an assembly plant and the Wheel Bogey assembly is supplied by a vendor.

Supply Chain Bill of Material for RY1OO - Part number 2000, Wheel Bogey is made by the supplier

The following diagram shows how the planning system in the assembly plant uses the master schedule and the bill of material for RY1OO to calculate the dependent demand for the 2000 Wheel Bogey. They then convert this demand into a vendor schedule that is transmitted to the supplying plant. The supplying plant then uses this customer schedule to develop their master schedule for the Wheel Bogey and calculates the requirements for its component.

Plant-to-Plant Linkage with Intelligent Demand

At first glance, the demand for the Wheel Bogey in the supplying plant looks like independent demand. It came from the customer and was not generated through the supplying plant's schedules and bills of material. However, if we look at the total supply chain, the demand for the Wheel Bogey in the supplying plant is directly linked to its schedule in the assembly plant. In fact, there is a "Supply Chain Bill" or a "Bill of Plants" that is creating this link. If this link is working properly, the demand for the Wheel Bogey in the supplying plant is dependent on the master schedule in the assembly or customer plant. We are using the term intelligent demand to refer to this special type of dependent demand driven by the schedule link between the two plants.

Now, put it into S&OP!

At DBM, we break the Sales and Operations Plan into five sections: Bookings, Shipments, Backlog (Open Customer Orders), Supply(Production) and Inventory.

We have found this presentation works in multiple business models, including build-to-stock, build backlog? Customers may provide a d projects. Using this standard format has facilitated combining multiple business models in a single S&OP process. However, the intelligent demand model provides some unique challenges, especially for the booking and backlog sections of the plan.

Using this definition, the backlog for this demand stream is the shipping plan over this schedule zone. In the example, we see a backlog of 120 units in Month 1,120 units in Month 2 and 110 units in Month 3. The total backlog is 350 units. backlog.

To address these issues, we have to establish a specific schedule zone that we will treat as backlog. This is a 'gutfeel' exercise. In the following example, the customer is providing a shipping plan for 12 months but we are only treating three months of this as actual backlog. We know that they may change the plan inside this zone but we are also confident that we must plan to meet this demand. There are a number of factors that go into setting this zone such as our specific customer contract, the supply chain lead-time, the demonstrated stability of the schedule, etc. In the end, the actual length of this zone is not as important as setting it. The difference between two months and three months will not dramatically change the results. I have defined the last period of this zone as EOB - end of backlog.

Using this definition, the backlog for this demand stream is the shipping plan over this schedule zone. In the example, we see a backlog of 120 units in Month 1,120 units in Month 2 and 110 units in Month 3. The total backlog is 350units.

With a shipment plan and an opening backlog, we can back into the bookings. The actual bookings for Month 0, the month just completed, are calculated as follows:

Actual Bookings= A-8

A = The sum of the Current Shipment Plan for M0 to EOB


B = The sum of the Last Shipment Plan from M0 to (EOB-1)

Let's put this in layman terms. A is the sum of last month's shipments an d the new shipment plan out to the end of the backlog (EOB) and B is last shipment plan from the beginning of last month to one period prior to the end of the backlog, actually the end of the backlog, as it would have been in the last plan. Basically, the actual bookings equal the shipments, plus or minus any changes in the backlog zone, plus the shipment plan for the new month added to the backlog.

So what? Why do we care about a booking and backlog plan in this case? Why go through all the trouble?

Tracking bookings gives us an early warning of changes in the demand stream. In the example, we see the Month 2 bookings dropping off to 80 units, foreshadowing the Month 5 shipment plan. You could get to the same story by looking at the shipment plan if there is only one demand stream. However, once you start adding multiple demand streams with different booking and backlog characteristics, the story gets lost.

Let's add another demand stream and then look at the aggregate demand picture for this family. This is a standard build-to-order demand stream with a target lead-time of six weeks. In this case, we would expect to see backlog to cover all of Period 1and half of Period2 Looking at the sheet, we see that there are 10 units past due and 100 units due in Month 1,

making a total of 110 units in the backlog for Month 1. There are 50 for Month 2, which is about right if the target lead-time is six weeks.

Now, if we add both demand streams together, we see the total demand plan that would represent the top three sect ions of the 5-Sect ion Sheet.

The aggregated demand shows that we are booked so lid for Month 1and have 60 units that need to be booked and shipped for Month 2 .From this presentation, we can see that we should protect space in Month 2 to maintain the target lead­ time for the build-to-order customers If we were only looking at the shipment plan, and not the backlog, this wouldn't be clear.

Most of the organizations we work with have multiple demand types for each of their families. To get to the story, they need a consistent format for each of the demand streams that are aggregated into the family plan. With this approach, they are able to treat the intelligent demand streams in the same manner as they treat their other types of demand and thus develop a clear and logical statement of total l family demand.

If you want to chat about this with me, feel free to contact me at dmcleod@dbmsys.com look forward to hearing from you!

Intelligent Demand Article
Download PDF • 8.49MB

At DBM, we provide your team with coaching and the tools needed to run an effective S&OP process. Through the Virtual Accelerator process, you can implement a sustainable Executive S&OP process in just six months.

Not quite sure how to incorporate these best practices into your S&OP process? We can help! Reach out to us. We can help you unlock the power of S&OP in your organization.