Forecasting Best Practices

Forecasting "Best Practices"

"Effective demand planning and sales forecasting across the supply chain can bring a host of benefits. Specifically, it can help improve labor productivity, reduce head count, cut inventories, and speed up production flows, and increase revenues and profits.

-Edward J. Marien

To find the "best practices" for forecasting, our team researched many cases of forecasting success, and found five companies with a common theme. Rayovac, the Coca-Cola Bottling Company, AAi. FosterGrant, the Sara Lee Corporation, and the Scotts Company all had major problems with forecasting, some of them very similar. To address and solve these problems each of these companies made major improvements to their forecasting systems. Although some used similar methods and others very different, these companies found that the right people, process and information technology was the key to efficient and accurate forecasting.

Rayovac

Rayovac discovered the proper use of forecasting when it implemented its consensus decision-making teams to improve their performance. They believe that the best information comes directly from the customer. Therefore, their forecasting takes place from the bottom-up. Sales, marketing, finance, and supply chain builds a consensus for forecasting, inventory management, scheduling, warehousing, and transportation. Their consensus forecasting is the responsibility of a cross-functional team.

For their annual planning process, a bottom-up and a top-down forecasting procedure is used. The top Key Results Measures (KRMs) are compared with the bottom-up forecast which is developed by the consensus teams. Meetings are held to align the two when there are discrepancies. Monthly meetings are held to update the bottom-up forecast and actions are taken to bring the forecast and annual plan into alignment.

Rayovac has discovered some key guidelines to be used for successful forecasting and planning:

1. Actually using the forecast to drive the business and not just to report accuracy.

2. The people must be committed to one process.

3. A forecast "champion" who has the support of upper management should be identified.

4. The right tools must be utilized, including forecasting software, supply chain software, and communication software.

They have also developed a monthly process with specific steps to be followed each week:

Week 1: Forecasts are used to build and maintain a quantitative forecast, a consensus forecast, and an annual operating plan based on KRMs.

Week 2: The information is reviewed by sales and supply chain management . SKU level information (unique promotions, new customers, etc) is added.

Week 3: Consensus team meetings are held.

Week 4: The forecast is disaggregated to distributor levels and is pushed to the ERP system. Inventory and production schedules are optimized.

Rayovac has seen improvements in communication among all functional areas, a combination of technology with human intuition and expertise. They also have shared responsibility, goals and risks. Their forecasts are now integrated and reconciled.

Coca-Cola Bottling Co. Consolidated

At Coca-Cola Bottling Co. Consolidated (CCBCC) has developed a forecasting system to change their previously decentralized forecasting and production processes. Their new centralized forecast allows the company to have more visibility throughout all facilities and the company can now be more proactive.

Their process involves marketing and customer-development groups feeding in price plans from the CCBCC's largest customers. Estimated demand, known promotions, and price points, new product introductions, and other inputs are added. They then forecast out 17 weeks, with attention focused on the next four weeks. The forecast is updated weekly and sent to area sales managers who review it and modify it as needed four weeks out. This allows a personal aspect to the technological system.

Allowing the forecast to change as things happen has led to improved communication and collaboration with common carriers. Also, inventory is reduced by about half and forecast accuracy has improved 10-20%.

Sara Lee

Sara Lee has also seen improvements thanks to an updated forecasting system. Professional demand planners and updated software tools has allowed them to progress. Previously, marketing had control of sales and financial forecasts and little attention was paid to them. Therefore, high amounts of safety stock and high amounts of stockouts ensued.

By developing a demand-planning group Sara Lee was able to get the appropriate people resources needed for good forecasting. They also were careful not to involve new technology that would disrupt the process already in place. With the new technology and demand planners in place, the company was able to get inputs from product managers and field sales people. With their incoming data automatically integrated into the demand planning software, they can forecast from the SKU level up to the family level. Field sales people are now being asked to forecast for the next four months for the largest customer because they have immediate access to the people and therefore, the best view of the marketplace.

Forecast accuracy is up approximately 51%, and perfect order fill rates are in around 93%. Continuous improvements are now much easier to attain as they now have good metrics and can see current trends.

Scotts

In 1999, the Scotts Company was dominating the lawn and garden industry. The company was already leading the U.S. market, and after making a number of major acquisitions including Ortho brand, the company had now became the number one supplier of lawn and garden products in Europe. But the company still had major problems in its supply chain, due largely in part to inaccurate and unreliable forecasts. Scotts "had no idea which items were actually needed, or where," and to make up for this "safety stocks were unacceptably high" (1). For a company with severely seasonal sales, and a highly complex supply chain with over 800 SKUs, the forecasting system in place was just not working. Forecasting was done on a national

...