Sales forecasting is the process of selecting and targeting specific market segments, and identifying which sub-markets are both viable and attractive requires the sales forecast or estimating future demand. Without a sales forecast, the marketing executive could not determine what sales volume to expect, how much to produce, and how much money and effort to spend on the entire marketing program.
A firm’s marketing efforts are affected by the world, national, and local economies. Because economic factors can greatly impact a firm’s sales, marketing executives may develop their forecasts through a top-down approach. Starting at the national or possibly world level, the analysis moves down to the local area and then to the industry.
Ultimately, the forecasts are narrowed to the firm’s products or services, as shown in the following figure :
While a top-down approach provides a good analysis of conditions affecting a firm’s sales, it is sometimes criticized because forecasts are built on forecasts, and estimates are made of estimates. The results can be a gross error in the firm’s sales forecasts if the preceding forecasts are not accurate.
To avoid this problem, some marketing executives prefer the bottom-up approach. Usually, using salespeople who are closest to the buying public and more knowledgeable about the people in their sales areas, sales estimates are made for each local area. These estimates are then combined into one overall forecast.
National and local levels
A marketing executive will frequently begin sales forecasting by examining the national economy. While very large businesses often have their own forecasts for the national economy, most other firms do not. Instead, they use forecasts prepared and published by the government, banks, and other institutions.
Once the national economic forecasts have been made or obtained from some other source, projections are needed for specific geographic areas to market its product or service. Although it is increasingly common for businesses to have their own local forecasts, most still use those developed by local governments and other sources.
Industry and company levels
Based on the general estimates of national and local economic conditions, the marketing executive then forecasts sales for the industry and the company. The industry forecast is synonymous with the firm’s product or service’s market potential, reflecting the total expected sales volume for the firm and all its direct competitors.
In some instances, industry forecasts are obtained from trade associations, suppliers, government agencies, and privately published sources.
The marketing executive may develop these industry forecasts by using the techniques described next. The company sales forecast, of course, is more important. The firm’s operations – including its production facilities, personnel, and financial plans – are geared to that level of sales.
Sales Forecasting Methods
Forecasting methods range from the judgmental to the highly quantitative, from the fairly simple and inexpensive to the quite costly and time-consuming. In selecting one or more methods to use, the marketing executive must consider the time and money available and the level of accuracy desired. The common methods are discussed below.
This is one of the most commonly used methods for forecasting sales. The marketing executive attempts to identify a pattern in sales volume by using past sales data. The pattern may be cyclical, constant, or consistent in movement, as illustrated in the following exhibit.
In a cyclical pattern, sales have recurring peaks and valleys based on a time dimension, like months or seasons of the year. A constant pattern, on the other hand, remains unchanged over the course of a time frame. Finally, in a consistent pattern, sales follow some predictable path.
Trend analysis is advantageous because it is simple and inexpensive to use. In some instances, it can be sufficiently accurate, especially in stable industries.
However, a simple trend analysis usually does not provide highly accurate forecasts. Seldom in the past, especially good pretors of the future because environmental variables change frequently and significantly, and the firms marketing efforts certainly do not remain constant.
For reasonably stable and consistent trends, a least square line is used, resulting simply in a trend line that is statistically closest to all of the past data used.
Cyclical sales require a number of techniques, such as weighted moving averages and exponential smoothing.
A weighted moving average typically places more emphasis on the recent past than earlier time periods. A more accurate and elaborate extension on this is exponential smoothing, which uses more quantitatively developed weights than the marketing executive’s subjective ones.
These techniques can be found in any basic mathematics or statistics book. There are other statistical techniques also available there that can help make trend analyses.
Market Factor Analysis
The market factor method is another technique commonly used for sales forecasting. The marketing executive finds another variable – usually another product – that has related sales patterns and for which the executive believes accurate sales forecasts have already been developed. This other sales estimate is the basis for the product’s sales forecast.
This method has the same advantages as trend analysis. If there is a logical relationship between the market factor and the firm’s product, the forecast will be quite accurate. Unfortunately, however, the relationships are sometimes thin at best and subject to sudden change.
A more sophisticated version of the market factor method is correlation analysis.
Instead of using judgment, several variables are statistically tested to closely determine how closely they correspond to the firm’s sales volume. The procedures can be found in any basic statistics book.
The actual equation used to make the sales forecast derived from analyzing past data on the variables and product sales. Weights, or coefficients, are attached to the variables, and a regression equation is then developed.
Correlation analysis is advantageous because it measures the accuracy of the regression equation created more precisely. Hence, the marketing executive knows how accurate it is likely to b in advance of a forecast. This method also is relatively easy and inexpensive to use.
One of the weaknesses of correlation analysis is the considerable amounts of historical data needed to make the comparisons. Usually, twenty past periods are considered necessary.
It is sometimes difficult to obtain sales figures that far back, making it impossible to use correlation analysis for new and near-new products. Moreover, it relies on past trends that are questionable, especially in rapidly changing items.
The trend analysis, market factor, and correlation methods all rely on past data and variables to derive sales forecasts for a firm’s product. Marketing executives often prefer to survey consumers directly in their target market to determine what and how much they will purchase.
Instead of guessing about people’s buying patterns for an upcoming period, why not just ask them?
Despite its attractiveness, however, a customer survey often does not provide the marketing executive’s information.
Asking customers what they will buy actually only measures their intentions, and intentions are frequently very different from actions during the course of a year. Many buyers do not know how much they plan to purchase, and others will not participate in surveys.
A test market is perhaps a more realistic approach to including customer actions in the sales forecast. Instead of measuring buyer intentions, the marketing executive sells the product in one or more limited areas and monitors the sales volume.
These test results are then projected to all market areas. Depending on the type of test market used, the results will indicate who the buyers are, how much they purchase, and what strategies competitors can be expected to use. Test markets are most commonly used to forecast sales for new products where no past sales figures are available.
Despite the usefulness of test markets, they do contain several distinct disadvantages. They are expensive to conduct and must be completed in a relatively short time.
Moreover, first-time purchases are not always good future sales indicators because buyers may try a product but not like it. Another disadvantage of a test market is that it exposes the product and the firm’s marketing strategies to the scrutiny of competitors.
Finally, it is sometimes difficult to find a good test market for the product. The area should be reasonably well populated, have the appropriate overall demographic profile of a target market, have good communication facilities, and consist of average users of the firm’s product.
Some marketing executives prefer using executive judgments instead of scientifically-based methods of developing sales forecasts.
Supporters of this method contend that top-level executives frequently have a feel for the firm’s markets and customers, a skill developed through years of experience and knowledge of the firm and its competitive environment. They further argue that there can be no substitute for the seasoned judgment of top-caliber experts.
This method’s primary advantages are that it is somewhat simple and inexpensive. But it may involve more than just a few quick questions to key personnel.
When properly done, executives hold a number of meetings after extensive individual study of market conditions. However, all too often, executives do not take the time to make a careful, informed judgment but rather a spur-of-the-moment guess.
A variation of the executive judgment method is the sales forecast composite, where the firm’s sales personnel make the forecasts. This approach’s rationale is that the salespeople are the firm’s closest link to the marketplace and its target market.
The salespeople are asked to estimate sales for their particular territories, and these are then added together, forming the overall sales forecast. This process is actually the bottom-up approach mentioned earlier.
One of the principal advantages of this method is that the forecast is made by people who are attuned to the marketplace every day. Probably the greatest dangers of this method are its lack of scientific base and its possible bias.
Using Multiple Methods of Sales Forecasting
Each of the methods described here can be used to forecast the sales for a firm’s product or service. Each has some unique strengths and weaknesses rather.
To take advantage of the strengths and nullify some of the weaknesses, marketing executives sometimes use several forecasting methods at the same time. In this way, checks and balances are developed, ensuring that the firm does not use an offbeat forecast.
In some cases, the firm will use an average of forecasts. In others, a single method will be used as the primary one, and several others used for verification purposes.
Overall, sales forecasting methods should be as simple and inexpensive as possible. Marketing executives must experiment to find the one(s) that best suits their needs and one in which the top executives have confidence.