Customer Model in Harmonized Planning Area

The current time-series-based supply and order-based applications offer different support for supply chains and specific customer models.

Time series and order-based planning have slightly different supply chains. Order-based planning supports individual sales orders on the ERP level (sold-to, ship-to, and partner, which is a business partner in ERP). Time series supports demands on a higher level, which is a virtual grouping of business partners, as shown in the diagram below:

The customer master data model provides three customers that represent the different dimensions that are required for the corresponding applications: time-series-based planning represents the planning dimension, order-based planning represents the operational dimension of the receiving customer, and inventory optimization represents the demand streams

• The customer (master data type TSP Customer) has TSP Cust. ID (I_TSPCUSTID) as its key attribute, which connects it to a location through a transportation lane (master data type TSP Customer Transportation Lane 
• The operational dimension of the receiving customer is modelled as follows: The customer (master data type Customer) – key attribute OBP Customer ID (I_OBPCUSTID) – is not connected to a location through a transportation lane. Instead, the connection between the customer in order-based planning and a location is established in the forecast or sales order.
• Demand streams are modelled as follows: A demand stream is represented by the Customer Group (I_CUSTGROUP) attribute, which is one of the key attributes of the Customer Group Source (I_SOURCECUSTGROUP)) master data type, and connected to a location through another key attribute of this master data type: Location ID (I_LOCID). The Customer Group attribute is also part of the TSP Customer master data type. The attribute values can be identical, but they don't have to be.

In particular, Time Series supply can support a customer model that accounts for the lead time between customers and plants, thereby modelling customer lead time and a customer sourcing ratio.

The time series forecast models the quantity as it arrives at the customer. And the order-based forecast models the quantity as it leaves the plant, so there's a slight difference between the two forecasts (time series and order-based), but the functionality existing today will be supported.

In the figure below, we can see what the customer model will look like.

The OBP customer is on the left of the figure. This reflects the sold-to/ship-to business partner. It's the existing customer, in order-based planning as of today.

Time series, in turn, will also have a customer object. It will be named the TSP customer.

TSP customers can be viewed as a grouping on a higher level, and time series can plan forecasts on a higher level than OBP.

To establish a relationship between the OBP and the time-series customer, a mapping can be introduced during supply chain modelling in implementation projects.

Some examples of project-specific configuration:

1. Simple Complexity
   1. Scale: a smaller BU
   2. Sample industries: B2B, Services
   3. Use case: selling out of a small number of locations to a small number of customers. Virtualization of customers is not needed; you can plan demand in a time series for real customers (sold-tos).
   4. The copy from TSP to OBP doesn’t require disaggregation; it's a matter of mapping one attribute to another.
2. Medium Complexity
   1. Scale: significant number of customers (10k+)
   2. Sample industry: Equipment Distribution
   3. Use case: customer groups defined for segmentation purposes (S&OP business discussion, IO service level management, etc.). Virtualization of customers is unavoidable, but the most important customers may still be forecasted by themselves (e.g. Amazon).
   4. Configuration can consist of direct mapping and disaggregation.
3. High Complexity
   1. Scale: significant number of customers (10k+)
   2. Sample industry: CPG/Food, Life Sciences
   3. Use case: customer groups segmented with an advanced technique on an aggregated level (like a state/province). Potentially complex packaging, sourcing, shelf life, regional regulations, etc.
   4. From the configuration standpoint, the direct mapping and disaggregation are difficult.
      1. In this case, it is important to assess OBP's expectations.
      2. What is integrated from SAP IBP to S/4 SAPHANA? (example: transport requisition, planned orders, allocations, etc.)
      3. Can the "weighting" of OBP customers be fine-tuned in so

The time series will continue to include a customer product, where the time series forecasts reside, and a customer product lane, which can model the customer lead time and the sourcing decision (customer sourcing).

To emphasize the differences between time series and order-based forecasts, consider that the primary focus of time series planning is on the customer product level, so the initial time series forecast, usually the consensus demand copied from demand planning, is an input on the customer product level to planning.

In order-based planning, the forecast is considered at the location-product level. There can be additional levels and more details about the plant level that ships to customers. Still, it's always location and product, and there's a difference between these two forecasts.

1. The first one is I_WKPRODLOCTSPCUST, which is today's demand planning and demand review planning level.

   This is where forecasting happens.

2. Then, the primary demands will be passed to the time series supply. This will be passed on to supply planning as forecast input. It will be at the TSP customer and category (CATID) levels in the model.

   The time series will take its forecast on the product customer combination, enhanced by category to distinguish different forecast streams.

3. Then time series supply can roll forecast demands up to the first plant location. These key figures are on the level of I_PRODLOCTSPCUSTMOT, which is the customer lane planning level.

   This is where the lead time shift and customer sourcing happen.

4. Finally, OBP runs forecasting or takes forecast as the primary demand on the level of I_PRODLOC. A new planning level called I_DAYPRODLOCDEMAND was introduced.
   1. Like in today's order-based planning area, it can be configured or left on product location.
   2. There is an option to enhance it with attributes like OBP Customer or Delivery Priority or any fields from the SFC catalog.