Wilkins, A Zurn Company: Demand Forecasting - Case Study Example

Wilkins Regulator Company is a manufacturer of high-quality products for plumbing, municipal waterworks, fire production, and irrigation customer markets.  Their products include water pressure reducing valves, backflow preventers, and anti-scald shower valves.   The market demands for Wilkin’s products are directed by new home starts, remodeling, and construction activities.  External factors that affect the business include the weather, unemployment rate, and the availability of financing.

Despite the different internal and external factors as well as the cyclical nature of the US commercial and industrial construction market that affect the revenues generated, the company has experienced sales growth that exceeds the industry. Their positive growth is due to their favorable pricing strategies, product innovations, and targeted marketing programs.The current demand forecasting process of the company includes a forecast master and a planning bill. The forecast master is a spreadsheet that lists the average weekly sales history for each product family by quarter and year since 1999.

For each product family, the total quarter’s actual sales are divided by 13 weeks per quarter to determine the average weekly sales per quarter. The expected demand for the next five or six quarters is then estimated for the quarters.Each product family had its own planning bill. It contains the sales history for each product within the family. It calculates the average number of units sold within that product family each day within each quarter. It also contains a projection on the average daily sales for that family that will sell in the next 12 months.

Furthermore, the planning bill disaggregates the family forecast into each product based on the percent of sales of the product family. Lastly, the planning bill calculates the annual sales forecast for each product within the family.The current forecasting performance utilized by the company is inaccurate. According to sales records, there are variances in the forecasted sales value with respect to the actual sales of the company. This paper will center on the utilization of statistical forecasting methods in order to improve the performance of the business.

The current method of forecasting that is utilized by the company does not take into account the seasonality of the sales generated. The forecast is based on the forecast master and the planning bill for each product family. The method of using two forecasting tools to estimate future demand is applicable to the company, however, the approach in estimating is inaccurate.  The company uses the naïve approach by estimating the expected demand for the product in future sales.  As a result, the forecast error is high.

  (See Appendix A) This paper utilizes statistical methods in forecasting by applying trend projection through linear regression as well as incorporating a seasonal index in the demand forecast in order to fit future demands with the trend of the historical sales.  The linear regression uses available actual sales data per quarter that would be used for the forecast master.  (See Appendix B) The idea behind the linear regression method is to illustrate a line that best fits the actual data, thus, a trend projection is created.

  The percentage variation in the value of the linear regression with respect to the actual sales value becomes the seasonal index.  The role of the forecaster is to identify the cyclical or seasonal trend in order to estimate the possible sales forecast. The figure below shows that the actual quantity sold from PVB is seasonal.  The linear regression shows that the trend is increasing.  By incorporating the seasonal index calculated from the linear regression forecast, the demand for the next quarter is estimated using actual figures and not through a naïve approach.

Figure 1.  DEMAND FORECAST USING LINEAR REGRESSION AND SEASONAL INDEXConclusion The implementation of statistical methods if forecasting minimizes human error in estimating the future demands of the product.  The naïve approach is based on the experience and gut-feel of the forecaster.  By employing trend projection as well as seasonal index analysis, the Wilkins Regulator Company can standardize its forecasting process as well as simplify the process without much interference from human intuition or “guesstimate”, such that, the results of the demand forecast will be founded on sound analytical sense and scientific methodology.