How Many Are We Going to Sell? Who Knows?

by Dave Garwood

After several decades of pursuing a better forecast, are we making any progress? The answer is yes. Lessons learned from the quality revolution and wireless internet access have added some helpful tools.The breakthrough came when the strategy shifted from searching for a formula to massage past history and project future demand to a more holistic approach of demand management. Demand management shifts the emphasis to looking to the future. It is important to understand the differences between forecasting and demand management ...

Forecasts have traditionally been derived from a mathematical calculation based on historical data. Demand planning takes much more into consideration. Future events such as promotions, price increases, new products and competitive situations are considered. The assumptions about these demand drivers are documented. Individuals clearly own segments or demand streams and are accountable for establishing the plan and meeting it. For example, the demand may be for 200 per month, but the southeast sales manager knows how much his team is accountable for selling. Action plans to make sure the plan is executed are established. The actual sales are frequently compared to plan and adjusted, if needed. The same demand data is used for driving all activities, including financial plans, in the business. Demand planning is viewed as a process. Quality replaced accuracy as the demand measurement and is used to drive continuous improvement. In short, a well-defined process to anticipate customer orders has been established.

Customer Order or Forecast -- What's the Difference?
A forecast is an anticipated customer order -- a commitment that hasn't yet been made official by a customer. A customer order is a forecast that has come true. Which one do we need to plan supply to support? Both! If we wait until after the customer places the order, we usually don't have time to make the product and meet the customer's delivery request or replenish finished goods inventory before we get another order. In short, a demand plan is needed if we have any chance of keeping the customers happy.

How do we plan future demand for a product? Consider three types of demands:

1. Customer orders: demands already inhouse and entered or booked.
2. Forecasted orders: potential customer orders that are being pursued by the sales people. Some of these may be a firm quotation or just an expression of interest by a potential customer.
3. Forecast: future demands that are anticipated, but we do not have a particular customer in mind yet.

The first one is the best demand prediction we will ever get. The second, forecasted orders, gives us a little less certainty but at least we have a specific target on the radar screen. Forecasts are usually necessary because it takes a lot longer to buy material and make the product (a parameter called Time Fence) than the customer is willing to wait for delivery or even think about committing to a product. As we look forward to anticipate future demand, all three types of demand must be considered. Look at the example below:

Several months ago, we did not have any customer orders or forecasted orders for December through May. The only demand was a forecast. We anticipated the demand during December through March to be 800 and had not looked beyond March. While we knew demand varies each month, we chose to use an average of 200 per month until we knew differently. Thanks to the heroic efforts of our crack sales team (and possibly a round of golf, dinner and a ball game) we now have some firm orders and a several hot prospects. In the meantime, the Supply Siders, i.e. manufacturing and purchasing, are waiting for a clear signal to react if the demand has changed or should they still plan on 200 per month? The sooner they know, the more likely they can react cost effectively.

We have a firm order for 125 in December and two additional hot prospects for 75 and 25 each. How hot are they? Is it likely both orders will be firmed up and the demand will be 225 for December? Are the orders part of or in addition to the total four-month forecast? What about January? We already have orders totaling 175 and four hot prospects totaling another 200. Should we just wait and see? If the demand turns out to be 375, we will have unhappy customers. If the hot prospects weren't that hot and we don't adjust demand now, we will have some excess inventory. Or should we just let the Supply Siders decide? Bad choice, after all, who is supposed to be closest to the customer? What about February and March? Are the forecasted orders in January part of the next two months forecast? Are we going out of business in April and May? The forecast currently drops to zero. How much demand should we plan in those two months? All good and critical questions ...

Here is the irony of this typical situation: someone will answer these questions -- often by default and the wrong person. Some planner or buyer will decide to not schedule more material for January delivery. In effect saying the forecasted demands won't happen. Or they got burned in the past when the demand did happen and start treating every order as in addition to the forecast, thus overstating the demand. The visible testimonial to this default action may be the layer of dust on products in the warehouse.

Demand Planning: A Critical Role in Supply Chain Management
Every business has a combination of the these three demands. Someone needs to determine the total demand plan to effectively manage the supply chain. This role is a demand planning function. Here are a few key questions that must answered as part of demand planning process:

• Are the forecasted orders going to happen? Do any of them need to be rescheduled? If customers expect a 4-week delivery, any forecasted orders in the next 4 weeks should likely be rescheduled out.
• Are the Firm and Forecast orders part of or in addition to the forecast?
• Does demand on this product affect demand on other products, i.e. does it cannibalize other product demands?
• Does the total demand pass a sanity check? For example, if we sold between 100 and 300 per month last year, does a demand plan of 500 next month make sense? Does a declining demand plan for a new product make sense?
• How does the total demand compare to the Annual Operating Plan?

High quality demand planning is no longer a luxury. It is a necessity. Many companies are using TQM principles and tools to improve the quality of the demand planning process. Wireless, handheld PDAs connected to the internet offer an effective way for sales people to continuously communicate forecasted orders.

These are exciting times. Problems once considered impossible to solve are being solved. Better forecasting is one of them!