Document View

Time series analysis, regression analysis and other techniques have become standard approaches to forecasting. However, management often lacks the expertise to judge forecasts. This problem can be somewhat overcome by relying on a need-performance (N-P) map. The N-P map is a tool that quickly tells executives how the company is doing. The N-P mapping requires customer surveys. A random sample from a customer database is chosen, and customers are asked to rate their needs for each attribute. The rating scale is simple. Customers compare an attribute as relatively high, medium or low in terms of need. Next, customers are asked to rank the company's performance in delivering the attribute. The survey procedure is sensitive to customer opinion in 2 ways: 1. It allows customers a chance to voice their opinions concerning company performance. 2. It categorizes the performance ratings relative to customer needs.

Business forecasting has long relied on quantitative tools to obtain best estimates of the future. Often, forecasters use sophisticated models to develop forecasts. Time series analysis, regression analysis and other techniques have become standard approaches to forecasting. But the management often lacks the expertise to judge forecasts. This problem can be somewhat overcome by relying on a need-performance (N-P) map.

Forecasting in a competitive market is based on the likelihood that customers will buy a product or service. Whether or not they would depends on how sound the company's marketing management is. By matching high business performance to high customer needs, the company can assure right products to right customers, high customer loyalty, reduced customer churn, and strong and continuous revenue streams.

How needs and performance can be best matched is illustrated with the help of a N-P map like the one shown in Figure 1. (Figure 1 omitted.) By periodically surveying its customers, companies can map customer opinion and determine whether or not business performance meets customer needs. The forecast of sales volume will be the highest if the company meets high needs with high performance as shown by the high/high zone of the map.

If the forecast is not what the management is looking for, it can improve it by re-prioritizing its activities and improving its operations. Excellent customer retention and gaining new customers are the rewards for hitting the high/high target.

JUDGMENTAL FORECASTING

The N-P mapping provides a qualitative forecast. With this, one can at least determine the general direction of a forecast. It is not in anyway a substitute for model based quantitative forecasts. Instead, it provides a basis to judge whether or not the quantitative forecast is appropriate.

Here the forecasting approach is based upon sampling opinion how effectively the company is meeting the expectations of its customers. The N-P map is a tool that quickly tells executives how the company is doing. For example, the company might be delivering high quality service where customer need is low. This paints a picture of customers' apathy and a situation where slow growth would be expected. Another example may be of a situation where the company experiences low performance in a high need area. This depicts an inferior growth outlook because customers are minimally satisfied. In a worse case, low performance might appear at the same time customer need is low. This portrays an extremely poor business situation, no hope of improving the sales of the company.

SURVEYING ATTRIBUTES

Business performance is clearly understood when a firm identifies specific attributes that are important to customers. For example, the attributes that are most critical to telecommunications customers, displayed on the N-P map, are:

a. Functional Utility

b. Product Quality

c. Product Features

d. Sales/Customer Service Professionalism

e. Pricing/Value

The N-P mapping requires customer surveys. A random sample from a customer data base is chosen, and customers are asked to rate their needs for each attribute. The rating scale is simple. Customers compare an attribute as relatively high, medium or low in terms of need. Next, customers are asked to rank the company's performance in delivering the attribute. Again, a simple scale of high, medium and low is used.

The survey procedure is sensitive to customer opinion in two ways: (1) It allows customers a chance to voice their opinion concerning company performance. (2) It categorizes the performance ratings relative to customer needs. Chart 2 provides an example of survey data from one customer. (Chart 2 omitted.) According to the map, business performance is excellent as far as product quality (b) is concerned. The high need attribute is being met with high performance. However, the customer has a serious problem with product features (c). This is the only attribute rated high in need, but it is rated low in performance. This means that the company does not meet the customer's expectations in delivering product features that are important.

COMPILING DATA

One customer's opinion is usually not sufficient for forecasting. A representative sample of customers is needed. Chart 3-A shows an example where one-hundred customers are queried about functional utility. (Chart 3-A omitted.) It is easy to see from the relatively symmetrical results that the average customer believes the attribute is medium in need and the company's performance is also medium.

Unfortunately, survey data are often unwieldy and more difficult to interpret. Chart 3-B shows a more realistic situation where the raw numbers are uneven, making it difficult to locate the average response by inspecting the chart. (Chart 3-B omitted.) In that case the weighted average can be used where weights are intuitively assigned. One can get a good reading about the customer need and performance by assigning a weight of 3 to the attribute of "high," 2 to the attribute of "medium" and 1 to the attribute of "low." The weighted average of attributes are plotted in Chart 3-C. (Chart 3-C omitted.) It shows that the customer need is moderately high (weighted average = 2.58) but the cmpany performance is above medium (weighted average = 2.45). So, here the company performance is somewhat lower than the customer need.

FORECASTING IMPLICATIONS

Turning to the example of telecommunications product, Chart 4 displays the central locations of five different attributes. (Chart 4 omitted.) In addition, X is plotted to show the overall central location of all attributes combined. The implicit assumption made in plotting X is that each attribute is equal in its effect.

The N-P map indicates that the comliany should expect neither a high forecast nor a growth level slightly higher than a moderate one. According to the survey, customers believe that only one attribute, product quality (b), is high in performance. Despite the excellent performance, customers' need is only medium. Because of this attribute customers cannot be expected to buy this product in large quantity. The need is not large enough to expect a high forecast.

There are two attributes that have high need--functional utility (a) and product features (c). Unfortunately, the company's performance is only medium. Therefore, a high forecast cannot be anticipated unless the firm initiates some improvements.

Reviewing all the attributes on the N-P map, the management has an excellent sense of potential forecast. The management is in a better position to review the analyst's quantitative forecast. Suppose the analyst used a model that projected high growth. The management could review the analyst's assumptions and determine whether or not they are consistent with the N-P results. For example, the analyst's high forecast might have been the result of a favorable pricing scenario. However, the N-P map demonstrates that price/value is only of medium importance. Such questions should arise and eventually be resolved, so that a reasonable forecast can be arrived at. In essence, business forecasts should not rely on models alone. Customers' disposition toward a product or service must be clearly understood. One can gain insight about the future sales of a product by relating the attitude ofcustomers with the performance of a firm.

Mr. Cook is a management consultant specializing in business forecasting and marekt analysis. Since 1976, he has assisted many large companies in developing models and managing forecasting process.