Methods and Techniques of Sales Forecasting

Sales forecasting methods and techniques vary from company to company. Every company that uses sales forecasts possesses its own technique to approach the forecasting process. Some companies have a dedicated team of forecast professionals while others use the sales staff to generate the forecast. The statistical methods used to generate the sales forecast depend on the demand profile of the product. Statistical forecast methods vary widely and finding the right method often boils down to trial and error.

Sponsored Link

QuickBooks From Intuit
Intuit QuickBooks is now in India. A Must Software for SMEs. Try Free! quickbooks.in

Decomposition

Decomposition stands as one of the most common statistical sales forecasting methods. Decomposition belongs to the time series family of forecasting methods. Decomposition looks at four variables that control the value of “x” over a certain time period. In simpler terms, decomposition uses a product’s trend component, cyclical component, seasonality and irregular components to forecast the future value of the product over a given time period. Decomposition looks at each component separately to determine a forecast value for the specified component and then combines the data output into an overall forecasted value. A variety of statistical decomposition methods exist.

Simple Exponential Smoothing

Unlike decomposition, which uses the entire history of a product as the forecast input, simple exponential smoothing uses a weighted moving average. Because simple exponential smoothing seeks to reduce, or smooth out, the irregular patterns in a product over time, this forecasting method works best with products whose main component exhibits strong cyclical and irregular patterns.

Census X-11

Census X-11 resembles a standard decomposition method because it uses the same variables trend, seasonality, cyclicality and irregularity as forecast inputs. The difference comes from how it uses these variables. It places more emphasis on the seasonal and cyclical components of the product. Census X-11 also uses a specific number of trading days in the month. Using trading days allows this forecast method to weigh the future forecast by the number of trading days used in the forecast input.

Techniques

All forecasters use a different technique when performing forecasting activities. Some forecasters prefer to forecast in a vacuum—not using input from other sources other than the data. This technique rarely works for any extended period of time. In most businesses the best source of data comes from the human elements involved in the business. Forecasting in a vacuum disregards this important source of data. Collaborative forecasting techniques such as collaborative planning, forecasting and replenishment use internal company resources and resources from suppliers to create a mutually agreed upon forecast.

Considerations

When considering what forecast methods and techniques to use remember this, forecasts are always wrong. The best method and technique is the one that keeps the business running efficiently and at the least cost. Don’t get bogged down in theories and methodologies or getting a 100 percent accurate forecast.

Abstract

Sales forecast plays a prominent role in business strategy for generating revenue. Sales forecast depends on some of the factors as the market demand, promotion strategy used, living standard of the people, inflation rate, consumables price, public image of the company, market share, quality of service and so on. In this paper sales forecast of Maruti Suzuki Ltd, an automobile industry in India is considered. The inflation rate, petrol price, previous month sale are found to be more prominent parameters influencing the sales forecast of cars in this company. The model is trained using Fuzzy Neural Back Propagation Algorithm. The result thus obtained is compared with other statistical technique like multiple regression technique. However the result obtained by proposed algorithm is found to be superior to the result obtained by multiple linear regression technique.

The sales forecast is a prediction of a business’s unit and dollars sales for some future period of time, up to several years or more. These forecasts are generally based primarily on recent sales trends, competitive developments, and economic trends in the industry, region, and/or nation in which the organization conducts business. Sales forecasting is management’s primary tool for predicting the volume of attainable sales. Therefore, the whole budget process hinges on an accurate, timely sales forecast. These technical projections of likely customer demand for specific products, goods, or services for a specific company within a specific time horizon are made in conjunction with basic marketing principles. For example, sales forecasts are often viewed within the context of total market potential, which can be understood as a projection of total potential sales for all companies. Market potential relates to the total capacity of the market to absorb the entire output of a specific industry.

On the other hand, sales potential is the ability of the market to absorb or purchase the output from a single firm. Many agencies and organizations publish indexes of market potential. They base their findings on extensive research and analysis of certain relationships that exist among basic economic data—for example, the location of potential consumers by age, education, and income for products that demonstrate a high correlation between those variables and the purchase of specific products. This information allows analysts to calculate the market potential for consumer or industrial goods. Sales and Marketing Magazine publishes buying power indexes. Its commercial indexes combine estimates of population, income, and retail sales to derive composite indicators of consumer demand according to U.S. Census Bureau regions, by state, or by the bureau’s organized system of metropolitan areas.

The buying-power index (BPI) provides only a relative value which analysts adjust to determine the market potential for local areas. Forecasting methods and levels of sophistication vary greatly. Each portends to assess future events or situations which will impact either positively or negatively on a business’s efforts. Managers prepare forecasts to determine the type and level of demand for both current and potential new products. They consider a broad spectrum of data for indications of growing and profitable markets.

Forecasting, however, involves not only the collection and analysis of hard data, but also the application of business judgment in their interpretation and application. For example, forecasting requires business owners and managers to not only estimate expected units sold, but also to determine what the business’s production (materials, labor, equipment) costs will be to produce those items. Computer-aided sales forecasting has revolutionized this process. Advances in computer technology, information highways, and statistical and mathematical models provide almost every business with the ability to execute complex data analyses, thus reducing the risks and pitfalls prevalent in the past. These advances have made the process and costs of forecasting practical and affordable for small- and mid-sized businesses.

Factors in Sales Forecasting

Sales forecasts are conditional in that a company prepares the forecast prior to developing strategic and tactical plans. The forecast of sales potential may cause management to adjust some of its assumptions about production and marketing if the forecast indicates that: 1) current production capacity is inadequate or excessive, and 2) sales and marketing efforts need revisions. Management, therefore, has the opportunity to examine a series of alternate plans that propose changes in resource commitments (such as plant capacity, promotional programs, and market activities), changes in prices and/or changes in production scheduling. Through forecasting the company determines markets for products, plans corporate strategy, develops sales quotas, determines the number and allocation of salespeople, decides on distribution channels, prices products or services, analyzes products and product potential in different markets, decides on product features, determines profit and sales potential for different products, constructs advertising budgets, determines the potential benefits of sales promotion programs, decides on the use of various elements of the marketing mix, sets production volume and standards, chooses suppliers, defines financing needs, and determines inventory standards. For the forecasting to be accurate, managers need to consider all of the following factors: HISTORICAL PERSPECTIVE.

As a starting point, management analyzes previous sales experience by product lines, territories, classes of customers, and other relevant details. Management needs to consider a time line long enough to detect trends and patterns in the growth and the decline of dollar sales volume. This period is generally five to ten years. If the company’s experience with a particular product class is shorter, management will include discernible experience of like companies. The longer the view, the better management is able to detect patterns which follow cycles. Patterns which repeat themselves, no matter how erratically, are considered to be “normal,” while variations from these patterns are “deviant.” Some of these deviations may have resulted from significant societal developments that carried an impact that filtered all the way down to your business’s sales performance. Management may compensate for these abnormalities by adjusting the figures to reflect normal trends under normal conditions. BUSINESS COMPETENCE. The ability of a company to respond to the results of a sales forecast depends on its production capacity, marketing methods, financing, and leadership, and its ability to change each of these to maximize its profit potential.

MARKET POSITION. Forecasting also considers the competitive position of the company with respect to its market share; research and development; quality of service, pricing and financing policies; and public image. In addition, forecasters also evaluate the quality and quantity of the customer base to determine brand loyalty, response to promotional efforts, economic viability, and credit worthiness. GENERAL ECONOMIC CONDITIONS. Although consumer markets are often characterized as being increasingly susceptible to segmentation in recent years, the condition of the overall economy is still a primary determinant of general sales volume, even in many niche markets. Forecasters incorporate relevant data that correlate well or demonstrate a causal relationship with sales volume.

PRICE INDEX. If the prices for products have changed over the years, changes in dollar volume of sales may not correlate well with volume of units. At one point in time when demand is strong, a company raises its prices. At another time, a company may engage in discounting to draw down inventories. Therefore, accountants devise a price index for each year which compensates for price increases. By dividing the dollar volume by the price indexes, a company can track its “true” volume growth. This process is similar to an inflation index, which provides prices in constant dollars. As a result, management is able to compare the price-adjusted dollar sales volumes.

SECULAR TRENDS. The secular trend depicts: 1) general economic performance, or 2) the performance of the specific product for all companies. If a company’s trend line rises more rapidly than the secular trend line, a company would be experiencing a more rapid growth in the rate of sales. Conversely, if a company’s trend line is below the secular trend line, its performance is below the market’s average. Management also uses this type of comparison to evaluate and control annual performance. TREND VARIATIONS. Although the secular trend represents the average for the industry, it may not be “normal” for a particular company. The comparison of company trends to secular trends may indicate that the company is serving a specialized market, or that the company is not faring well. Forecasters study the underlying assumptions of trend variations to understand the important relationships in determining the volume of sales. Although markets may be strong, the sales force might need to be adjusted.

“INTRA-COMPANY” TRENDS. By analyzing month-to-month trends and seasonal variations over both the long and short terms, small business owners and managers can adjust the sales forecast to anticipate variations that historically repeat themselves during budget periods. Management may then construct a budget reflecting these variations, perhaps increasing volume discounts during traditionally slow periods, exploring new territories, or having sales representatives solicit product and service ideas from current customers. PRODUCT TRENDS. Forecasters also trend individual products, using indexes to adjust for seasonal fluctuations and price changes. Product trends are important for understanding the life cycle of a product. SOURCES AND MAGNITUDE OF PRODUCT DEMAND.

In past eras, the introduction of new and improved products drove much of the demand. Currently, consumer attitudes and lifestyles anticipate product introductions and technological changes. Individual consumers are pushing technology to anticipate the needs of an increasingly segmented market. Demand based on anticipation is becoming the dominant feature of the technological age. The rapid pace of technological development and new product introduction have shortened product life-cycles. The combination of demographic considerations and technological change dominate consumer trends to a greater degree than in the past.

Forecasting Techniques

There are a variety of forecasting techniques and methods from which the small business owner may choose. Not all of them are applicable in every situation. To allow for adequate forecasting, a business must choose those methods which best serve their purposes, utilize accurate and relevant data, and formulate honest assumptions appropriate to the market and product. Sales forecasts may be general if they calculate aggregate sales attainable in an industry. Conversely, forecasts may be very specific, detailing data by individual products, sales territories, types of customers, and so forth. In recent decades market analysts have increased their use of focus groups, individual surveys, interviews, and sophisticated analytical techniques aimed at identifying specific markets.

APPROACHES TO FORECASTING. In the causal approach forecasters identify the underlying variables that have a causal influence on future sales. The company has no influence over causal variables in the general society, such as population, gross national product, and general economic conditions. A company does, however, maintain control over its production lines, prices, advertising and marketing, and the size of its sales force. After studying the underlying causes and variables in depth, the analysts use a variety of mathematical techniques to project future trends. On the basis of these projections, management derives its sales forecast. The non-causal approach involves an in-depth analysis of historical sales patterns. Analysts plot these patterns in graphs in order to project future sales.

Because no attempt is made to identify and evaluate the underlying causal variables, the analysts assume that the underlying causes will continue to influence the future sales in the same manner as in the past. Although analysts may apply certain statistical techniques to extrapolate past sales into the future, this approach is sometimes criticized as simplistic or naive, especially since most business experts believe that rapid changes in technology are driving fundamental changes in many business operations. Analysts employ the indirect method by first projecting industry sales. From this data they project the company’s share of the industry total. The direct approach, however, skips the industry projection with a straightforward estimate of sales for the company. Either of these methods are applicable to the causal and non-causal approaches.

Forecasting Methodologies

A variety of sales forecasting methodologies can be used by small and large businesses alike: BOTTOM-UP FORECASTING. Analysts using this methodology divide the market into segments, and then separately calculate the demand in each segment. Typically, analysts use sales force composites, industry surveys, and intention-to-buy surveys to collect data. They aggregate the segments to arrive at a total sales forecast. Bottom-up forecasting may not be simple because of complications with the accuracy of the data submitted. The usefulness of the data is contingent upon honest and complete answers from customers, and on the importance and priority given to a survey by the sales staff.

TOP-DOWN FORECASTING. This is the method most widely used for industrial applications. Management first estimates the sales potential, then develops sales quotas, and finally constructs a sales forecast. Problems arise with this method, however, when the underlying assumptions of the past are no longer applicable. The correlation between economic variables and quantity demanded may change or weaken over time. These two forecasting methods encompass a number of methodologies which can be divided into three general categories: qualitative, times-series analysis and regression, and causal. QUALITATIVE METHODS. Qualitative methods rely on non-statistical methods of deriving a sales forecast. A company solicits the opinion or judgment of sales executives, a panel of experts, the sales force, the sales division supervisors, and/or outside expert consultants. Qualitative methods are judgmental composites of expected sales. These methods are often preferred in the following instances: 1) when the variables which influence consumer buying habits have changed; 2) when current data is not available; 3) when none of the qualitative methods work well in a specific situation; 4) when the planning horizon is too far for the standard quantitative methods; and 5) when the data has not yet factored in technological breakthroughs taking place or forthcoming.

The Probability Assessment Method (PAM) forecasts sales volume by utilizing in-house expert opinion that provides probabilities between one and 99 percent, plus and minus, on certain target volumes. Analysts translate these estimates into a cumulative probability curve by plotting the volumes by the probability assigned to them. They use this curve to aid in forecasting. The Program Evaluation and Review Technique (PERT) requires estimates of “optimistic,” “pessimistic,” and “most likely” future circumstances. Analysts weigh these three estimates to form an expected value from which they compute a standard deviation. In this way analysts convert the estimates of the small business owner and/or staff into measures of central tendency and dispersion. The standard deviation enables the forecaster to estimate a confidence interval around the expected value. While PERT is only an approximation, it is quick and easy to use. The forecaster can take into account the owner’s opinion as a check on estimates produced by other methods. The Delphi Technique relies on the assumption that several experts can arrive at a better forecast than one.

Users of this method solicit a panel consensus and reprocess the results through the panel until a very narrow, firm median is agreed upon. By keeping the panel participants isolated, the Delphi excludes many aspects of group behavior, such as social pressure, argumentation, and domination by a few members, from causing undue influence. The expense associated with this method, however, precludes most small business enterprises from pursuing it. A visionary forecast relies on the personal insights and judgment of a respected individual. Although often supplemented by data and facts about different scenarios of the future, the visionary forecast is characterized by subjective guesswork and imagination and is highly nonscientific. But while such forecasts are not based on reams of scientific information, many small business owners have achieved success by relying on such subjective data. Historical analogy methodologies, meanwhile, attempt to determine future sales through an in-depth analysis of the introduction and sales growth of a similar product.

Historical analogy seeks patterns applicable to the product considered for current introduction. This method requires several years’ history for one or more products, and—when used—is generally applied to new product introductions. The sales force composite gathers forecasts from each individual salesperson for a particular territory. The sales forecast is the aggregate of the individual forecasts. The usefulness of this method is dependent on the accuracy of the data submitted. An intention-to-buy survey measures a target market’s intention to buy within a specified future time period. Market analysts conduct such surveys prior to the introduction of a product or service. Analysts provide consumers with a description or explanation of the product or service with the hope that respondents will provide honest answers.

If respondents tell analysts “what they want to hear,” the survey will not be accurate. In addition, certain environmental factors, such as a competing technological breakthrough or a recession, may influence respondent buying habits between the time of the survey and the product introduction. TIME-SERIES ANALYSIS AND PROJECTION. Trend projection techniques may be most appropriate in situations where the forecaster is able to infer, from the past behavior of a variable, something about its future impact on sales. Forecasters look for trends that form identifiable patterns which recur with predictive frequency. Seasonal variations and cyclical patterns form more obvious trends, while random variables make projection more complex. While time-series methods do not explicitly account for causal relationships between a variable and other factors, analysts find the emergent historical patterns useful in making forecasts. Analysts typically use time series for new product forecasts, particularly in the intermediate and long-term. The data required varies with each technique.

A good rule of thumb is a minimum of five years’ annual data. A complete history is very helpful. Market research involves a systematic, formal, and conscious procedure for evoking and testing hypotheses about real markets. Analysts need at least two market research reports based on time series analyses of market variables, and a considerable collection of market data from questionnaires and surveys. In its simplest form, trend projection analysis involves the examination of what has happened in the past. Analysts develop a specific linear percentage trend with the expectation that the trend will continue. The problem with the simple trend projection is the fact of randomness—that is, the random event or element that has a major impact on the forecast. The moving average is a more sophisticated type of trend projection. It assumes the future will be an average of the past performance rather than following a specific linear percentage trend.

The moving average minimizes the impact of randomness on individual forecasts since it is an average of several values rather than a simple linear projection. The moving average equation basically sums up the sales in a number of past periods and divides by the number of periods. Industry surveys involve surveying the various companies that make up the industry for a particular item. They may include users or manufacturers. The industry survey method that uses a top-down approach of forecasting has some of the same advantages and disadvantages as the executive opinion and sales force composites. A regression analysis may be linear or multiple. With linear regression, analysts develop a relationship between sales and a single independent variable and use this relationship to forecast sales. With multiple regression, analysts examine relationships between sales and a number of independent variables. Usually the latter is accomplished with the help of a computer that helps analysts to estimate the values of the independent variables and to incorporate them into a multi-regression equation.

If analysts find a relationship among various independent variables, they can develop a multiple regression equation for predicting sales for the coming year. Exponential smoothing is a time-series approach similar to the moving average. Instead of using a constant set of weights for the observation made, analysts employ an exponentially increasing set of weights so that more recent values receive more weight than do older values. More sophisticated models incorporate various adjustments for such factors as trends and seasonal patterns. Analysts look at the leading indicators because the National Bureau of Economic Research has clearly demonstrated their value in forecasting. These indicators include prices of 500 common stocks, new orders for durable goods, an index of net business formation, corporate profits after taxes, industrial materials prices, and the change in consumer installment debt. Despite their widespread use, the leading indicators do not relate well with specific products.

Nevertheless, when such relationships can be established, analysts construct multiple regression models with which to forecast sales. CAUSAL METHODS. When analysts find a cause-effect relationship between a variable and sales, a causal model may provide better forecasts than those generated by other techniques. Life-cycle analysis forecasts new product growth rates based on analysts’ projections of the phases of product acceptance by various groups—innovators, early adapters, early majority, late majority, and laggards. Typically, this method is used to forecast new product sales. Analysts’ minimum data requirements are the annual sales of the product being considered or of a similar product. It is often necessary to do market surveys to establish the cause-effect relationships. The Sales Budget

The sales forecast provides the framework for the detailed planning presented in the master budget. Based on planned strategies and its best business judgment, management converts a sales forecast into a sales plan through the commitment of resources and the establishment of control mechanisms. The sales budget provides an evaluative tool by presenting monthly indexes of volume of units and dollars as hard targets for the sales team. Deviations from these indexes indicate to small business owners and managers where they need to adjust their efforts to take advantage of hot products or to remedy difficult situations. Management determines its sales policies and strategies within its ability to respond to customer needs, technological changes, and the financial prerequisites of marketing. The sales budget projects that portion of potential sales the sales team believes it can achieve.

The forecast, then, sets the parameters on the top side while the production capacity and sales acumen of the team sets the floor. Although sales forecasts may accurately project significant changes in market conditions, a company needs to thoroughly examine its own resources to determine its ability to respond to these changes. A huge drop in demand may decrease the strain on the production process to where a company regains cost efficiencies, or a large increase in demand might be required by a company that needs cash for other projects. The sales budget, therefore, is predicated on a company’s ability to meet expected demand at or near its maximum profit potential. THE PRODUCTION BUDGET. Both small and large businesses construct their production budgets within limitations of production, warehousing, delivery, and service capabilities. Subsequently, a company attempts to schedule production at maximum efficiency.

By anticipating the variations in monthly sales, management can keep production at levels sufficient to provide adequate supply. Labor costs generally comprise the greatest single production cost. Therefore, management may adjust labor hours to production schedules. Production levels remain rather constant if current inventory is sufficient to meet increases in sales. If management expects an increase, it may build inventories during the first quarter of the budget, and sell them down to planned levels during the remaining three quarters. From the production budget, a company estimates the mix of materials, labor, and production overhead needed to meet planned production levels.

Developing a Sales Forecast

Forecasting sales is inherently more difficult than the construction of the subsequent sales budget. Although management exerts some degree of control over expenditures, it has little ability to direct the buying habits of individuals. The level of sales depends of the vagaries of the marketplace. Nonetheless, a sales forecast must attain a reasonable degree of reliability to be useful. Fundamentally, sales forecasters follows steps similar to these in developing a forecast of sales potential: * Determine the purposes for using the forecasts.

* Divide the company’s products into homogeneous groups. * Determine those factors affecting the sales of each product group and their relative importance. * Choose a forecasting method or methods best suited for the job. * Gather all necessary and available data.

* Analyze the data. * Check and cross-check deductions resulting from the analyses. * Make assumptions regarding effects of the various factors that cannot be measured or forecast. * Convert deductions and assumptions into specific product and territorial forecasts and quotas. * Apply forecasts to company operations.

* Periodically review performance and revise forecasts.
Further Reading:

Bolt, Gordon J. Market and Sales Forecasting. Franklin Watts, 1988. Cohen, William A. The Practice of Marketing Management. Macmillan Publishing, 1988. Crosby, John V. Cycles, Trends, and Turning Points: Practical Marketing and Sales Forecasting Techniques. NTC Publishing, 2000. Henry, Porter, and Joseph A. Callanan. Sales Management and Motivation. Franklin Watts, 1987. McCarthy, E. Jerome, and William D. Perreault, Jr. Basic Marketing: A Managerial Approach. Irwin, 1990. Mentzer, John T., and Carol C. Bienstock. Sales Forecasting Management: Understanding the Techniques, Systems, and Management of the Sales Forecasting Process. Sage, 1998.

Read more: http://www.answers.com/topic/sales-forecast#ixzz2bZlypNp9