Introduction to demand forecasting

This is a presentation covering the concepts of demand forecasting. it includes the meaning of demand forecasting, purpose, scope and factors affecting demand forecasting. It also covers the methods of forecasting for both new and existing products.

1. Demand forecasting Prepared by- Amanda Bvera

2. MEANING OF DEMAND FORECASTING Demand results in sales which constitutes the primary source of revenue for the business. A forecast is a prediction or estimation of a future situation, under given conditions. Demand forecasting predicts the future trend of sales given the present state of demand determinants. Demand forecasting is different from demand estimation in the sense that the former predicts about the future trend of sales while the latter tries to find out expected present sales level, given the present state of demand determinants. Forecast can be both physical as well as financial in nature, and are used mostly for planning purposes.

3. PURPOSE OF DEMAND FORECASTING Forecasting is done for both the long run as well as short run. The purpose of the two, however, are different. In a short run forecast seasonal patterns are of prime importance. Such a forecast helps in preparing suitable sales policy and proper scheduling of output in order to avoid over-stocking or costly delays in meeting the orders. Besides giving an idea of likely demand, short-run forecasts also help in arriving at suitable prices for the products and in deciding about necessary modifications in advertising and sales techniques. Long-run forecasts are helpful in proper capital planning. When installing production capacity, an element of flexibility in their availability has to be ensured to take care of planned and expected changes in production. Long-term planning thus helps in saving the wastage in material, manhours, machine-time and capacity. Long-run forecasting is usually used for ‘new unit’ planning, expansion of the existing units, planning long-run financial requirements and man-power requirements.

4. DETERMINING THE SCOPE OF A FORECASTING EXERCISE Before taking up a forecasting exercise, one has to know the specific purpose for which the predictions are sought. The sales forecasts can have differing interpretations and applications, depending upon the purpose of the forecast. Defining scope of the forecasting exercise is, therefore, a necessary pre-requisite. There are at least 6 considerations which need to be taken care of while determining the scope of a forecasting exercise: Period of forecast Level of forecast General purpose or specific purpose forecasts Forecasts of established products or new products Type of commodity for which forecast is to be undertaken Miscellaneous factors to be included or not.

5. Period of forecast: As a first step one has to decide about the length of the period for which the forecasting exercise is taken up. The time periods are usually divided into (a) Short-run (b) medium term (c) long-run. Short-run forecasting: In case of short-run forecasting, one is looking for the factors which bring fluctuations in the demand pattern in the market. The most important factor in this regard maybe the weather conditions. As the weather conditions influence the demand for consumer goods, it indirectly affects the demand for machinery, equipment, raw material, etc., needed to produce these consumer goods. Thus, seasonal factors are the ingredients of short-run forecasts. Medium-term forecasting: In case of medium-term forecasting, experience and sound judgment are more important than statistical forecasting. The medium-term forecasts can assist in the decision about timing of an activity, like advertising expenditure. These forecast also contribute to control or revision of the decisions based on long-run forecasts. The main feature of medium-term forecasts is the trend. The direction of the trend, has important implications for subjects like employee recruitment and training, etc. Long-run forecasting: In the long-run, the validity of the trend itself must be ascertained. If the trend is likely to change for the worse, this would have implications for the firm’s entire long-term strategy, it could suggest the wisdom of diversification policy for the firm. For long-run forecasts, reliance is usually placed on statistical techniques, though judgement still remains an asset in identifying the variables which are mainly controllable and are likely to influence future sales. In the very long-run analysis one must include variables relating to socio-psychological aspects-in particular an analysis of the inter-relationship of economic, psychological and sociological factors determining consumer behaviour.

6. Levels of forecast: Sales forecasting may be undertaken at any one of the following levels: Macro-economic forecasting: It is concerned with business conditions over the whole economy. These business conditions are measured with the help of some appropriate indices like those relating to national income, industrial production, wholesale prices, etc. These indices are provided by official and non-official agencies and these can be treated as basic assumptions on which to base the demand forecasts. Industry demand forecasting: Such forecasts can give indications to a firm regarding the direction in which the whole industry will be moving. These forecasts, based on surveys of consumer’s intentions and analysis of statistical trends, are generally supplied by trade associations to its members. The business may use such forecasts to compare the trends of industry sales with its own current and expected sales and thus determine its market share. Firm demand forecasting: A big firm will like to do forecasting of its own products independent of the rest of the firms in the industry. Such forecasting reveals whether or not the company is well placed to maintain or even better its share in the market. Product line forecasting: It helps to decide which of the product or products should have priority in the allocation of the firm’s limited resources.

7. General purpose or specific purpose forecasts: Though a general forecast is useful for a firm, it will be even more helpful if the general forecast is broken down into specific forecasts with respect to commodities, area of sale, domestic and export markets. Forecasts of established or new products: Problems and methods of forecasting differ in these two cases . For the established products past sale trends and competitive conditions are known, while this is not the case for new products, hence suitable forecasting methods need to be adopted in this regard. Type of commodity for which forecast is to be undertaken: Economists broadly classify good into three categories – capital goods, consumer durables and non-durable goods. For each of these categories of goods there would be distinctive demand patterns. Mis cellaneous factors to be included or not: The forecaster should decide how much the sociological and psychological factors are going to enter into the forecasting exercise. To be more effective, this exercise must also include the factors like the features peculiar to the product and the market, nature of competition, impact of uncertainty and risk and the consequent errors in accuracy, etc.

8. STEPS INVOLED IN DEMAND FORECASTING The following steps are necessary to have an efficient forecast of demand: Step 1: Identification of objective Step 2: Determining the nature of goods under consideration Step 3: Selecting a proper method of forecasting Step 4: Interpretation of results This can be explained as follows: Identification of objective: It is necessary to be clear about what does one want to get from the forecast. The purpose of the exercise may be the estimation of one or more than one aspect, like quantity and composition of demand, price to be quoted, sales planning, inventory control, etc. The approach to the problem will accordingly differ.

9. Determining the nature of goods under consideration: Different categories of goods have their own distinctive demand patterns. It is therefore, necessary to determine the class in which the goods fall. These categories are: the capital goods, consumer durables and non-durables. This will help us in identifying the approach of the forecasting exercise and in determining the variables to be considered for forecasting. Selecting a proper method of forecasting: The selection of an appropriate method of forecasting is related to the objective of forecasting, type of data available, period for which the forecast is to be made, etc. For example, if the data shows cyclical fluctuations, the use of linear trend will not be suitable. Similarly, general trend may be more useful for long-run forecasting, while seasonal patterns will be more important for the short-term forecasts. Interpretation of results: Mere preparation of forecast does not lead the management anywhere. Interpretation of results is equal importance. Efficiency of a forecast depends, to a large extent, upon the efficiency in the interpretation of its results. Most of the times, the forecasted results are to be well-supported by the background factors which have not entered the exercise of forecasting. Further, we need to frequently revise the forecasts in the light of changing circumstances because forecasts are, in the first instance, made on the assumption of continuation of past events.

10. SIGNIFICANCE OF DEMAND FORECASTS The significance of demand forecasting is shown in the following points: Fulfilling objectives Preparing the budget Stabilizing employment and production Expanding organisations Taking managerial decisions Evaluating performance Helping the government This is explained in the following manner:

11. Fulfilling objectives: Implies that every business unit starts with certain pre-decided objectives. Demand forecasting helps in fulfilling these objectives. An organisation estimates the current demand for its products and services in the market and moves forward to achieve the set goals. Preparing the budget: Plays a crucial role in making the budget by estimating costs and expected revenues. Stabilising employment and production: Helps a firm to control its production and recruitment activities. Producing according to the forecasted demand of products helps to reduce the wastage of the resources of the firm. This further helps the organisation to hire human resources as per the requirement. Expanding organisations: Implies that demand forecasting helps in deciding about the expansion of the business. If the expected demand for products is higher, then the organisation may plan to expand further. On the other hand if the demand for products is expected to fall, the organisation may cut down the investment in the business. Taking managerial decisions: Helps in making critical decisions, such as decided the plant capacity, determining the requirements of raw materials and ensuring the availability of labour and capital. Evaluating performance: Helps in making corrective decisions, for example if the demand for the firm’s products are low, it may take corrective actions to improve the level of demand by enhancing the quality of its products or spending more on publicity and advertising. Helps the government: It enables the government to coordinate import and export activities and plan international trade. It could also pertain to boosting the level of production in the country.

12. DETERMINANTS OF DEMAND FORECASTING Goods can broadly be classified into three categories as follows: Durable consumer goods Non-durable consumer goods Capital goods While forecasting demand for each of these categories, we keep in mind different considerations because different types of goods have their own distinctive demand patterns. This is explained in detail in the following slides.

13. Consumer durables: Each consumer durable has a special market for its product, which in turn has its peculiarities. So, forecasting demand for individual products in such cases requires special techniques adapted to meet these peculiarities. Moreover, the demand for consumer durables falls into two categories: replacement demand and new demand. Forecasts should be made separately for both. The special difficulties or peculiarities in forecasting in case of consumer durables are as follows: Changes in size and characteristics of population. Saturation limit of the market Existing stock of the goods Replacement demand vs. New demand Income level of the consumers Consumer credit outstanding Tastes and scales of preferences of consumers The forecaster of consumer durables, therefore has to use to different techniques to predict his future level of sales.

14. Non-durable consumer goods: These include those consumer goods which can only be used once, e.g., food, beverages, tobacco, etc. Demand for such goods is basically influenced by the following factors: purchasing power (or disposable income) of the consumers (Yd), price of the commodity (P) as well as population and its characteristics (S). Symbolically, it is D = f(Yd, P, S) Capital goods or producer’s goods: Capital goods are defined as those goods which help in further production of goods. Capital goods include factory buildings, machinery, equipment, tools, etc. Capital goods are, therefore, demanded only when there is a demand for the goods which these capital goods help in producing. In other words, demand for a capital good is a derived demand, which will depend upon the profitability, level of capacity utilisation and wage rates in the industries using the capital good. Moreover, demand for a capital good is of two kinds: Replacement demand New demand

15. The following information is needed for estimating capital good’s demand: Growth possibilities of the industries using the capital goods. The norm of consumption of capital goods per unit of installed capacity. It is assumed that the norms of consumption often remain stable. However, in practise one finds cases where shortages arise, like in case of construction of bridges where mild steel is sometimes used when constructional steel is not available. Thus, though the norms usually remain stable but they are sometimes violated too. The extent of excess capacity in the industry using the producer’s goods. The forecast of demand for the good which the producer’s goods help in producing. The existing stock of the producer’s goods. The age-distribution of the existing stock of producer’s goods The rate of obsolescence. The availability of funds and costs of funds to the firms using the producer’s good. The nature of tax provisions on re-equipment. The prices of substitutes and complementary goods. The market structure within which the producer’s good operates.

16. Methods of demand forecasting The methods of forecasting can be divided into two categories namely, opinion polling methods and mechanical methods. The opinion polling methods deals with the collection and aggregation of various units- their demand patterns, consumption patterns, spending plans, etc. These methods are subjective, more expensive and can also be time consuming. These methods are useful for predicting the demand for new products and also for predicting the short run demand forecasts. Mechanical methods are cheaper, faster and easier to calculate as they are based on the records already available with the firm.

17. Consumer survey: In this method, the consumers are contacted personally and are invited to share and disclose their consumption patterns and demand preferences. This method is advantageous in the sense that it collects first hand information and is free form other biases. This method is conducted in three forms: Complete enumeration: Here, the entire population is surveyed and the results are collected and aggregated. Sample survey: Here only a sample or a unit from the relevant population is surveyed and the results from various such units are collected and aggregated. End-use: In this method, the input-output values are used. The firms using the particular good as an intermediate good are surveyed and hence the results obtained from this procedure are aggregated.

18. Sales force opinion method: This method is called the reaction survey method in which the opinion of the salesmen, that is people closest to the market are surveyed and their responses are collected and aggregated. An advantage of this method is that it is cheap and easy, in the sense that it does not involve any elaborate statistical measurement. Another advantage to this method is that it is based on the first-hand knowledge of the salesmen. Expert’s opinion method: Obtaining views from a group of specialists outside the firm has the possible advantages of speed and cheapness. This method is best suited in situations where intractable changes are occurring. It is possible in cases where basic data are lacking experts may give divergent views, but even then it is possible for the manager to adapt his thinking on the basis of these views.

19. Time series: The time series analysis lays on the use of time as a variable in the analysis of seasonal patterns and trends in the sales of the products. The variables analysed under this method are: Trend: It is the overall movement of the time series. Cyclical fluctuations: It is the wavelike movements and intermittent changes in the time series. Seasonal patterns: It is the seasonal up and down movement in the time series that occur at a particular time. Irregular movements: These are the movements left after the isolation of the other three variables.

20. Barometric series: Under the barometric method it is assumed that there is some economic relationship between the various time series. Based on this, there are three types of relationships among the series as follows: Leading series: It is moving ahead of the time series it is compared to. Coincident series: Moving along the series it is compared to. Lagging series: It is moving behind the series it is compared to. Forecasting through barometric techniques involves the following stages: To locate the leading indicator for the variable whose forecast is being undertaken. To estimate a mathematical/statistical relationship of leading indicator with the variable under forecast. To find out the forecasted values of the variable. If possible, to verify the validity of the forecast with the help of coincident indicators and/or lagging indicators.

21. Correlation and Regression method: One of the popular methods of forecasting. It realises the fact that more variables other than time affect the value of sales. The numerous variables influencing sales are identified through correlation and analysed under regression with the help of a model, to determine the future trend of sales. Econometric method: The difference between the barometric and econometric method is the scope of variables that are analysed. It analyses a vast number of economic and demographic variables and builds up a cause-effect relationship. Forecasting through econometric models involves three stages: Identification of variables and the functional form. Estimation of parameters. Finding the forecast values.

22. Criteria for selecting a good forecasting method In modern times, particularly due to fast expanding communication and transportation facilities, the rate of change in demand has accelerated a great deal. Consequently, the problems of adjustment have also grown for a firm. There is an increased complexity of products and processes. The adaptability and flexibility of competing producers in adding new products and processes also lead to need for further adjustments. It is necessary for a firm, therefore, to know about the impact of future conditions. For this, it needs to undertake demand forecasting. There are many methods for this purpose. The final method of forecasting must be chosen on the basis of the following criteria: Accuracy Longevity Flexibility Acceptability and Simplicity Economy Availability

23. The results achieved by a forecasting method must be weighed against the cost of the method. But the results that can be achieved by a forecasting method are difficult to estimate before the cost of making the forecast has been expended. Hence the choice of method must be done on a priori grounds. The use to which the forecast can be made should be well understood. In fact, it is not a question of results achievable but that of results achieved by a forecasting method. For this, the persons making the decisions must fully understand the forecasting methods, their assumptions and probabilities. It is quite easy to judge the existing trend. But for a good forecast it is necessary that it should also predict deviations and turning points so that the forecasts are more effective. There is a time gap between the occurrence of an event and its forecast-Known as the ‘lead’ time. Longer the lead the forecast has before the event, the greater will be its usefulness. One may even sacrifice some accuracy for gaining a lead rather than sacrificing a lead for accuracy.

24. Demand forecasting for a new product To forecast the demand for new products, we can use either of the following four methods: Survey of buyer’s intensions or consumer survey Test marketing Life cycle segmentation analysis Bounding curving method Forecasting the demand for new products is always a challenge for the manager. As for the existing products the past trend of sales is already known. But for new products no such information is available. Hence, there is an element of risk in forecasting the demand for such products. Here, the sound judgement and calculative decision style

25. Consumer survey: In this method, consumers are contacted personally to disclose their future purchase plans. This may be attempted with the help of either a complete survey of all consumers called as complete enumeration, or by selecting a few consuming units out of the relevant population called as sample survey. In case the commodity under consideration is an intermediate product, then the industries using it as an end-product are surveyed known as end-use method. Test marketing: This method is a variation of the sample survey method. It involves selecting a test area which can be regarded as a truly representative portion of the total market. The product must then be launched in the test area in a manner identical to that which is intended to be used if and when the product is launched in the broad national market. Package design, sales force, TV and press support, price and so on, must be selected with this consideration in mind. If the product is successful in the test area, a forecast can then be made that similar levels of success will be achieved in the total market. Despite the fact that this method has the advantages of a real life experiment. It has many disadvantages too, such as: It is exceptionally costly in terms of both time and money. The test must be continued for a long period in order to allow the consumer repurchase cycle to occur, else false predictions about the success of the product would be made. It is difficult to select a test market which is representative of the total market. Rival producers may immediately imitate the product in the test market and launch it in the national market.

26. Life cycle segmentation analysis: Each product has a life cycle consisting of introduction, growth, maturity, saturation and decline. Since business tactics differ at each stage of the product life cycle, we would want to know when the product will be at any one of these stages. But since the total market also has its segments, life cycle for the same product may be proceeding at different rates in different market segments. Timing and the type of tactics will, therefore, differ according to the segment concerned. Conceptually, the five stages in the product life cycle and the relevant elements for their marketing mix are as follows: Introduction: Quality has the greatest marketing impact, then advertising; but price and service has the least impact. Growth: Early adopters have now already purchased the good; buyer resistance is now being met; advertising is the most effective weapon and then the quality. Maturity: Most price insensitive buyers have now bought; rivals have entered the market; so price elasticity has become very much higher. So price is most important, followed by advertising, quality and service. Saturation: Price is no longer important because it has already become low; product differentiation in quality, packaging or advertising become important. Decline: The problem now is to find new product uses and advertise them; quality and service will have some impact, but price very little impact. Bounding curves method: Bounding curves show the outermost limits of change in the market share which have occurred for existing brands. According to this method, taking the market share data of all the exiting brands, the firm will forecast the demand for its new product.

27. Thank you