Accurate demand forecasting is the first step in an effective sales and operations planning (S&OP) process. Over-forecasting or under-forecasting are equally problematic, and result in slow-moving or out-of-stock situations. All demand forecasts are in stock keeping units (SKU) and by stocking location.
5 tips for demand forecasting
-
Know what to include
Only independent demand is subject to demand forecasting. This includes only stocked items, subject to random demand for sale to the final customer.
Raw materials are not subject to demand forecasting, as the required quantities can be determined via the bill of materials. Non-stocked and buy-in items are also not subject to demand forecasting. This is one of the reasons for the importance of accurate master data.
-
Adapt the method and detail of the forecast
The basis of effective demand forecasting is adapting the method and detail of the demand forecast to what is appropriate for each stock keeping unit (SKU) category (“A”, “B”, “C”, etc.) in each stocking location.
The starting point of any demand forecast is based on actual historical demand. Ideally, demand should be used, but few organisations keep track of all demand, so actual sales is the usual input. This is calculated automatically by the computer system, and usually the level of sophistication of the algorithms is not as important as is assumed.
-
Get input from your sales and marketing team
Send the demand forecast that has been automatically generated by your computer system to sales and marketing for their comment and refinement. They will add or subtract from the automatic demand forecast based on their knowledge of the future and market. It’s easier to get people to refine an initial demand forecast than to give them a blank sheet and ask them to provide it from scratch.
-
Keep demand forecasts separate to targets
It’s also important to separate the demand forecast from the budget or target for that SKU and location. Sales forces are incentivised to meet and exceed their targets, but forecasting sales simply based on the difference between sales-to-date and the target results in a complete imbalance of stocks.
Demand forecasting is about trying to predict the market demand as accurately as possible. If sales are below target, this is a matter for increasing demand by promotions, advertising campaigns, etc.
-
Be clear about roles and responsibilities
A demand forecasting process must have clearly defined roles and responsibilities. Each sales and marketing team must forecast their own region, based on the warehouse footprint from which their customer demand is fulfilled. They must also have demand accuracy as a key element of their key performance indicators (KPIs) and a commission or bonus scheme.
Indicators of potential for improvement
Inventory management is a cornerstone of effective S&OP. Here are some typical symptoms of an inadequate inventory management process:
- Under-target customer service performance (e.g. customer back-orders and missed on time in full (OTIF) targets)
- Unbalanced stock holdings (e.g. out-of-stocks of “A” and an excess of slow moving or obsolete (SLOB) stock)
- Regular need to expedite supplier orders and/or expensive inbound freight options
- Excessive working capital tied up in stock and SLOB inventories
- Placement of replenishment orders on suppliers ad-hoc based on “specials” etc.
- The need for urgent customer orders and expensive outbound freight options
- General settings of safety stocks set in units and not specific per SKU category in sales cover