This is a dangerous assumption - markets and the economy are dynamic, not static - but from the practical point of view, there is little choice. In any case, it is usually over longer periods of time that changing market factors make themselves felt.

A distinction is necessary here between making a sales forecast and estimating sales potential. A sales forecast is based on past sales performance and a reckoning of known and anticipated market conditions.

From these, the expected sales level is determined.

Sales potential, on the other hand, is a measure of the capacity of the business to reach a certain volume of sales. It is based on knowledge of the total market and the extent of competitive influence, and it involves the use of strategy through sales effort. Past sales performance may bear little or no resemblance to sales potential. In general, sales potential is likely to represent a higher sales level than a sales forecast.

For the purposes of a buy-sell transaction, a short-term or at most an intermediate-term forecast is all that should be attempted. Short-term forecasts cover a few months - seldom more than a year. Intermediate-term forecasts should be limited to 1 or 2 years.

Since the forecast is based on past sales of the company, it is necessary to know the dollar sales volume of the firm for the past several years. If not enough sales data have been recorded, it may be necessary to improvise.

In one instance, the prospective buyer of a self-service laundry was unable to get sales figures. He contacted the manufacturer of the washing machines to determine the amount of water used per machine load. He then learned from the water company the amount of water consumed by the business. Using these two figures and making allowances for water used for drinking, rest room, and so on, he computed the number of loads washed per month. This figure multiplied by the price charged per load gave him a reasonably accurate figure for the sales volume.

For a short-term forecast, it is usually enough to know the sales for the past few weeks or months in comparison with the corresponding period of the year before. If sales for the past 4 weeks were 8 percent more than the corresponding 4 weeks of the preceding year, sales for the next few weeks can reasonably be expected to be 8 percent ahead of the corresponding period a year ago.

Adjustments have to be made, of course, for any known or predicted conditions that will change this rate of increase - conditions such as unusual weather, short-lived labor disputes, changes in the dates of events such as Easter, and so on.

Distribution of sales by months. A longer method of forecasting is based on the distribution of sales by months. This method works bast if the monthly variations over a period of years have been small.

Suppose, for instance, that a short-term forecast is being made in June. For the past several years, sales in July have been between 11 and 13 percent of annual sales, with an average of 12.5 percent. During the same period, May sales have averaged 10 percent of annual sales. Sales during the May just past were $16,000. Then $ 6,000 : 0.10 = $160,000, the estimated annual sales. Projected sales for July will be 12.5 percent of $160,000, or $20,000. Sales for other months can be forecast in the same way.

Cumulative percents. Another method of short-term forecasting is the cumulative-percent method. The percent of total sales is figured for each week during the past year and added to the percent for preceding weeks, as shown in this example :

Weeks        Weekly percent           Cumulative percent
1                       0.9                               0.9
2                       1.1                               2.0
3                       1.4                               3.4
4                       1.7                               5.1
5                       1.9                               7.0
6                       2.4                               9.4
7                       2.6                             12.0
8                       2.9                             14.9
9                       3.1                             18.0

If sales during the first 4 weeks amount to $8,000, the annual total will be estimated at $8,000 : 0.051, or $156,862. To forecast sales for the next 4 weeks, add the percentages for those weeks and multiply the annual estimate by the result ($156,862 X 0.098 = $15,372. This method works best for goods or services that are not subject to wide variations in sales volume and whose prices do not fluctuate greatly.

Number of sales transactions. Where prices tend to vary, the number of sales transactions may show a steadier trend than dollar sales do. An increase in dollar sales without an increase in the number of transactions means that the average dollar value per transaction has gone up. This increase in the amount of the average sale may mean (1) that customers are buying higher-quality goods, (2) that they are buying in larger quantities, or (3) that prices have increased.

If the level of transactions is steadier than the dollar sales, the forecast tends to be more conservative. A study of the transactions may bring to light factors not revealed by total dollar sales.

Because of the combination of variables at work in the market, the techniques used in the short-term forecast are not reliable when applied to the longer periods covered by intermediate-term forecasting. In the longer forecast, two methods of measurement are generally used: the long-term trend method and the correlation method. Correlation analysis requires data usually beyond the reach of the small businessman, but the long-term trend as determined by the least squares method may be useful. This method will not be taken up here, but an explanation of its use can be found in any introductory book on statistical methods.

It must be reemphasized that a trend is determined from past data and from the total market as reflected in company sales. Insofar as these conditions remain in about the same state of balance, a projection of the trend into the future has some value; but the more dynamic these market factors are, the less reliable trend lines become.

The investigator must give careful thought to how changing market factors will affect his forecast. Although he cannot have precise knowledge of these factors, he must decide how influential they are likely to be and adjust his forecast accordingly.