Managing Your Supply Chain Using Microsoft Axapta: A Book ExcerptPart One: Sales and Operations Planning

Managing Your Supply Chain Using Microsoft Axapta: A Book Excerpt
Part One: Sales and Operations Planning
Featured Author - Dr. Scott Hamilton - March 23, 2004

Introduction

A firm's sales and operations planning (S&OP) process starts with the definition of all demands for the firm's goods and services. It formulates game plans that drive supply chain activities to meet those demands. Hence, an effective S&OP game plan requires consideration of both demands and supplies. The nature of each product's game plan depends on the environment. The game plan may focus on stocked end-items in distribution and make-to-stock manufacturing environments. The game plan for a make-to-order manufactured product depends on the level of stocked components, the approach to defining product structure, and the need for direct linkage between production orders and sales orders. In this case, the game plan can be expressed as master schedules for stocked components and finishing (or final assembly) schedules for make-to-order items. Other considerations impact the nature of the game plans. A multisite environment may require consideration of inventory replenishment across a distribution network. Variations in operations such as projects and lean manufacturing also affect the game plan. The saleable products and services for many firms represent a mixture of environments.

Independent demands provide the logical starting point for formulating an S&OP game plan. The logic underlying planning calculations and demand-pull philosophies is built on chasing demands. Independent demands typically consist of sales orders or forecasts or a combination of both.

Since the nature of an item's S&OP game plan depends on the environment, several common scenarios are used to illustrate considerations about demand and how to formulate a game plan. The scenarios included here represent several types of distribution and manufacturing environments.

An effective S&OP game plan results in fewer stock outs, shorter delivery lead-times, higher on-time shipping percentages, a manageable amount of expediting, and improved customer service. Several guidelines are suggested to improve a firm's sales and operations planning process and the effectiveness of each product's game plan.

This is Part One of a four-part excerpt from the book Managing Your Supply Chain Using Microsoft Axapta by Dr. Scott Hamilton.

The book can be ordered on amazon.com.

Part Two will discuss "Understanding Planning Calculations".

Part Three will present "Common Scenarios".

Part Four will propose "Guidelines and Case Studies".

Reprinted with permission from McGraw-Hill

Identifying Demands

Sales orders identify actual demands. Actual demands drive all supply chain activities when visibility of the sales order backlog exceeds the cumulative lead-time to obtain and ship a product. Lead-time reduction efforts stemming from just-in-time philosophies can help companies produce to actual demand. However, actual demands must be anticipated in many situations such as selling a product from inventory or producing items from stocked components. One approach to stocking products involves a sales forecast, and the combination of sales forecast and sales orders defines demand for the saleable item. A second approach involves an order-point replenishment method where the minimum quantity point represents forecasted demand over the item's lead-time.

Sales Order Demand

A sales order line defines an actual demand for an item, expressed as a quantity, delivery date and ship-from warehouse. Another type of sales order—a subscription sales order for a recurring sale—also defines actual demand. In addition, a quotation sales order and CRM quotation can optionally represent a demand. A CRM quotation is considered a demand when its probability exceeds a specified percentage.

Sales Forecast Demand

A sales forecast defines an item's estimated demand, expressed as a quantity, date, ship-from warehouse, and forecast identifier. A sales forecast represents the desired inventory level on the specified date.

Each forecast entry for an item has a user-defined forecast identifier termed the forecast model. Using different forecast identifiers allows multiple sets of forecast data. Planning calculations are based on a specified set of forecast data. Multiple sets of forecast data often reflect various scenarios for simulation purposes, or forecast revisions based on changing market conditions. This approach supports comparison of actual demand to a selected set of forecasted demand. In addition, a sales forecast for an item can be associated with a specific customer or group of customers for comparison purposes.

Some companies use the concept of a two-level forecast model. With a two-level forecast model, for example, the forecast identifier representing the company-wide forecast can be associated with several forecast identifiers representing regional forecasts. Each forecast entry has a forecast identifier corresponding to a regional forecast, so that the system automatically rolls up the company-wide forecast.

Alternative Methods for Entering Forecast Data

Several alternative methods can be used as shortcuts for entering multiple forecasts for an item. These methods include a repeating pattern and a group of items, and the two methods can be used in combination.

Repeating Pattern. A repeating pattern (termed an allocation method) results in the automatic generation of quantity and date entries, either as fixed or variable quantities per period. The repeating pattern approach can be used for forecasting an individual item or a group of items.

- Fixed quantity per period. The user specifies a fixed quantity, a period (such as month), and a starting and ending date. The system automatically generates a periodic sales forecast for the fixed quantity across the specified time horizon.

- Variable quantity per period. This method employs a user-defined template (termed a period allocation key) that spreads out a total quantity across several periods based on a mix percentage per period. The period allocation key, for example, could have percentages assigned to each month in a twelve month horizon, such as 5 percent for January, 9 percent in February, and so forth. This approach can be used to model a seasonal or trend demand pattern.

Group of Items. A forecast can be defined for an item group rather than an individual item. This method employs a user-defined template (termed an item allocation key) that spreads out a total quantity across several items based on a mix percentage per item. The items must be in the same item group, and each template entry also defines the ship-from warehouse. For example, the template entries could define the mix percentages for shipping the same item from different warehouses.

Handling the Combination of Actual and Estimated Sales Demand

When using forecasted demand, the combination of sales orders and forecasts must be considered to avoid doubled-up requirements. Some companies manually maintain the sales forecast to correctly model the combined demand, where planning calculations do not employ any forecast consumption logic. However, the system supports two other approaches for automatically handling the combined demand. The approach to handling the combined demand is termed the forecast reduction principle (also known as forecast consumption logic) for planning calculation purposes. The three forecast consumption approaches are described below.

- No Forecast Consumption Logic. Planning calculations add the two demands (stemming from sales forecasts and sales orders) for determining requirements. This approach applies to environments with no forecasted demand, or with a manually maintained forecast that models the combined demand.

- Forecast Consumption by Open Orders. The forecast reduction principle of "open orders" means that sales orders consume forecasts within each forecast period. The forecast period provides a logical time span (such as monthly) for comparing actual sales orders to sales forecasts within the period, and making assumptions about the combined demands for an item. For simplicity's sake a monthly forecast period will be used for further explanation. In a given month, an item may have a single forecast or multiple forecasts; multiple forecasts typically indicate weekly, intermittent or even daily demands that drive supply chain activities. Any sales order line item with a delivery date within the month consumes the sales forecasts within the month, starting with the earliest forecast and consuming forward. The sales forecasts within a given month can be over consumed; there is no carry-forward effect to consume forecasts within a future period. As time moves forward, an item's sales forecast becomes past due when the system work date matches or exceeds the forecast date. Planning calculations ignore past due sales forecasts; there is no carry-forward effect within a forecast period.

- Forecast Consumption based on Reduction Percentages. The reduction percentage approach ignores actual sales orders for forecast consumption purposes. It automatically reduces forecasts based on a user-defined template (termed a reduction key) of a reduction percentage by forecast period. This approach reflects the concept of a demand fence where planning calculations ignore forecasted demands within a specified time horizon, and the time horizon represents the sold-out backlog for an item. Let's say an item always has a sold-out backlog of one month and a partially sold-out backlog in the second month, so that 100 percent of an item's sales forecast can be ignored for the first month and 50 percent for the second month relative to today's date. The user defines these percent reductions in the reduction key assigned to the item, and different keys reflect the nature of a sold-out backlog for different items.

The last two approaches to forecast consumption apply to stocked end-items, and to make-to-order standard products with stocked components. In the latter case, the backlog of sales orders drives the near-term production of make-to-order items while the combination of sales forecasts and sales orders drive the procurement and production of long lead-time items.

A forecast period can be daily or yearly rather than monthly. For example, a daily forecast period might be used in conjunction with a daily repeating pattern, and a sales order line item only consumes a day's forecast.

Statistical Forecasting and Demand Planner

The Demand Planner supports the development of a statistical forecast based on extracted data about an item's shipment history. It automatically determines the best fit with various statistical models to suggest a sales forecast. Projected quantities can be optionally overridden, and then used to automatically update a set of sales forecast data.

Purchase Forecast Demand

A purchase forecast defines an estimated demand for a stocked component, expressed as a quantity, date, ship-to warehouse, and forecast identifier. A purchase forecast is different from a sales forecast, and it requires consideration of forecast consumption logic to avoid doubled-up requirements. The three approaches to purchase forecast consumption are described below

- No Forecast Consumption Logic. This approach applies to environments with no purchase forecasts, or with a manually maintained forecast that models the combined demand.

- Forecast Consumption by Open Orders. The creation of a purchase order consumes the purchase forecast for a purchased item. In a similar fashion, the creation of a production order consumes the purchase forecast for a manufactured item. The previously described logic about the forecast period applies to a purchase forecast.

- Forecast Consumption based on Reduction Percentages. This approach automatically reduces purchase forecasts based on a user-defined template of reduction percentages, as described above.

Anticipating Demand Using Order Point Logic

An order-point replenishment method provides an alternative to forecasts when anticipating demand for a warehouse's stocked material. Time-phased order point logic (such as min/max) suggests replenishment when an item's inventory balance falls below its minimum quantity. The minimum quantity represents estimated demand over the item's lead-time. The system supports fixed or variable quantities for minimum and maximum quantities, where a variable quantity might be expressed in monthly periods. In addition, the minimum quantity can be automatically calculated based on historical usage.

Anticipating Demand Variations Using Safety Stock

Many firms carry additional inventory to anticipate variations in customer demand, and meet customer service objectives regarding stockouts, partial shipments, and delivery lead-times. The additional inventory is commonly called an inventory plan or safety stock. An inventory plan is typically expressed for items at the highest possible stocking level, such as saleable end-items that are purchased or manufactured to stock. A make-to-order manufacturer, on the other hand, typically expresses an inventory plan for stocked components.

An item's inventory plan can be explicitly expressed as a safety stock quantity when using a period or order-driven reorder policy. The system supports a fixed or variable safety stock quantity, and can calculate it based on historical usage. An inventory plan can also be expressed implicitly. An implicit inventory plan, for example, represents the extent to which a minimum order quantity exceeds typical demand over lead-time. An order quantity multiple can also inflate the order quantity so that it exceeds typical demand over the reorder cycle.

Other Sources of Demand

Visibility of all demands is critical to formulating an effective S&OP game plan. Surprise demands can cause shortages that impact customer service or production, and result in expediting. Some sources of demand may need interpretation or alternative ways to express the demand, as illustrated in the following examples.

Customer Schedules. Customer schedules represent a combination of sales orders (in the near-term) and forecast (in the longer term), and often require time-frame policies for proper interpretation.

Internal Sales Orders. An internal sales order may be required to initiate production or procurement activities prior to obtaining the customer's purchase order. Once obtained, the designated customer can be changed on the sales order.

Customer Service Demands. Customer service may require material for loaners, exhibition items, donations, replacement items and repairs.

Field Service Demands. Field service may require spare parts for selling to customers, and for repair and field service projects.

Engineering Prototypes. Prototypes may be built for internal or external customers, with requirements for material and production capacity. Procurement and production activity may also be initiated on new products with partially defined bills.

Quality. Quality often requires validation lots or first articles, especially during ramp up to production lot sizes. Other quality-related demands can be embedded in planned manufacturing scrap, so that planning calculations identify the additional requirements for material and capacity.

Purchase Returns to Vendor. Anticipated returns to vendor represent demands, as defined by return purchase orders that have not yet been posted as shipped.

Projects. The demands associated with internal and external projects can be forecasted.

This concludes Part One of a four-part excerpt from the book Managing Your Supply Chain Using Microsoft Axapta by Dr. Scott Hamilton.

The book can be ordered on amazon.com.

Part Two will discuss "Understanding Planning Calculations".

Part Three will present "Common Scenarios".

Part Four will propose "Guidelines and Case Studies".

Reprinted with permission from McGraw-Hill

About the Author

Dr. Scott Hamilton has specialized in information systems for manufacturing and distribution for three decades as a consultant, developer, user, and researcher. Hamilton has consulted worldwide with over a thousand firms, conducted several hundred executive seminars, and helped design several influential ERP packages. He previously co-authored the APICS CIRM textbook on How Information Systems Impact Organizational Strategy and recently authored Managing Your Supply Chain Using Microsoft Navision. Hamiliton is currently working closely with Microsoft partners involved with manufacturing and distribution, and can be reached at ScottHamiltonPhD@aol.com or 612-963-1163.