Forecasting Practices in the Electronic Commerce Retailing Industry

Article excerpt

ABSTRACT

This article reports findings from a study of retail E-Commerce firms' forecasting practices. The majority of such firms prepare sales forecasts, and top managers and financial managers most frequently prepare the forecasts. Most forecasts are for the entire company and are for annual time periods. Managerial judgment and percent of last year's revenue are the most common forecast techniques employed, and forecast accuracy varies considerably. The findings are discussed and recommendations are offered.

INTRODUCTION

Sales forecasts have been described as "Predictions of future revenues [that] are important to business planning, and are used to set production schedules, budget capital, and allocate resources to marketing programs" (Dalrymple, 1975, p. 69). Sales forecasting has been carried out, in one fashion or another, probably for as long as commerce has existed. Today, the managers of large and small concerns recognize this function's importance for all planning and control decisions. Despite the universal need, there is considerable variation across industries and firms in the manner the forecasting process is performed, and it is anything but a standardized process. Furthermore, sales forecasting practices change over time as the objectives and strategies of management and the dictates of the environment undergo alteration.

Retail electronic commerce (E-Commerce) firms require accurate and reliable sales forecasts to the same degree as any other type of business. Yet, due to considerable developmental changes, the E-Commerce industry has become synonymous with uncertainty (Golicic et al., 2001). Some "dot-com" firms have disappeared, and many that remain spark both curiosity and apprehension among investors. Further obstacles to effective forecasting in the retail E-Commerce industry include factors such as dramatic increases in online purchases, aggressive competition, the brief history of time series data, and a paucity of insights regarding on-line consumer behavior. These forces can hinder efforts to control and plan the operations of these firms so that resources can be optimally allocated and strategies formulated.

To date, most inquiries into forecasting practices have focused on developed industries like manufacturing, traditional (bricks and mortar) retailing, financial institutions, and the purveyors of services. This is a logical concentration, as these sectors customarily have encompassed the bulk of the commercial activity taking place. Yet, the increased activity and the degree of uncertainty in the E-Commerce industry confirm the need for an inquiry into the forecasting activities of E-Commerce retailers. In this article, we describe a study of forecasting activities in the retail E-Commerce industry. We report upon the degree of forecasting, who prepared the forecasts, what techniques were used, the time period and business unit, and the accuracy of the results. We also compare the retail E-Commerce figures with those of some traditional industries in an attempt to provide insights that may help academicians and practitioners understand sales forecasting performance differences between E-Commerce and traditional retailers.

THE INQUIRY

A questionnaire derived from past studies of forecasting practices (Dalrymple, 1967, 1975; Jun and Peterson, 1991; Peterson, 1993) was employed for our research. The questionnaire was adapted for the retail E-Commerce industry and was field-tested on a group of 50 retail firms randomly selected from the Yellow Pages in phone books from across the United States. Feedback from the respondents in the field-test was employed to further edit the questionnaire so that it was appropriate for the E-Commerce industry.

The sample for the main study was again drawn from Yellow Pages. Although this publication does not include all web sites, it contains a comprehensive listing of a cross section of Internet retailers, and is assumed to represent the more viable sites on the Internet. …