Forecasting Tools
Forecasting refers to a planning tool that assists the management to cope up with the hidden uncertainties that may happen in the future, it relies on the past and present data. Forecasting is started with certain beliefs, assumptions which are relying on the experience, knowledge of the management.
There are various techniques that are used to forecast the estimates such as the Delphi technique, top down and bottom up approach, projection, scenario planning, surveys of intention, model building, etc. The tools are been explained below:
The Delphi technique is a technique that is used for forecasting. It is also a method that takes use of the results obtained out of the questionnaires. It is also distributed to the panel of experts. It was been initially created as a systematic method of forecasting which relies on a panel of experts.
The experts are there to answer the questionnaires provided in rounds respectively. After each round the responses of the experts are shared among the group. It is further used to collect responses and collect data from the developed questionnaires on researched subjects (Hasson, F., & Keeney, 2011).
The Delphi technique is having various benefits such as, the experts doesn’t have to come together at a single place for meeting up. At the same time, when designing and implementing the Delphi technique, there are some tools such as the selection of subject, time of conducting and completing the study, the unintentional feedback of the respondents, etc which are needed to be considered.
The advantage of Delphi technique or method is that it helps to gather opinions from different set of experts and that too can be obtained without the need of gathering everyone for a physical meeting.
For example:
This technique is helpful and can be applied for saving the travelling expenses for a particular business firm.
The process of the Delphi technique is below:
Firstly, it begins with the open-ended questionnaires from a selected subject after that the investigator has to structure the feedback obtained by the respondents. Secondly, the investigator will hand-over the questionnaire by asking the respondents to review them carefully and then ranking the issues by order.
Thirdly, the respondents are expected to review the summary of the investigator from the last round of questionnaires and also provide the reasons of disagreement. Lastly, review is obtained from the last round by listing the items left, their ratings to the panelists (Rowe, G., & Wright, 2011).
Projection method is a classical method that is used for forecasting which deals with the variables movements through time. This method needs a long time-series data. When the numerical data is available a trend can be presented on a graph paper which shows changes through time.
This projection helps to know that where should be the trend at some point in the near future. The projection method relies on the belief that the factors will continue to maintain their role in future in the same manner as they did in the past.
These factors will determine the variable’s direction and also its magnitude (Yeh, te al., 2011). Though been a complicated method, Projection method is still one useful way to look at the past sales and also helps to discover the likely trends from the data and at the same time utilize the information to conclude what might happen in the near future.
The advantages of this method is that it is very easy to use, it provide accurate and appropriate forecasts, and is very quick and economical. But this method can only be used if the past data is available and also it is not mandatory that the trends of past will hold the same value in the future as well.
For example: A population that has a annual growth of 2% will double in about 20 years.
Top down and bottom up approach of forecasting:
Top down forecasting is a method or type of qualitative analysis which finds the themes related to investment and also the trends going through the help of sector/industry growth, the underlying opportunities of global regional, the macroeconomic picture and the geopolitical landscape.
The top down approach as the name signifies starts from the industry estimations and then works down to the individual departments, the products and the services.
The top down approach works really well when the forces in the market are similar on various products and sales areas. It works well for the budgets and strategy planning. On the other hand there is bottom up approach of forecasting.
The bottom up approach of forecasting is a quantitative analysis which identifies the possible growth prospects by the help of positive earnings, the reasonable valuations, and also the future expectations of growth.
It starts by gathering crucial or vital information from the individual products estimates and thus adding up to make a larger picture for the overall business. The bottom up approach is quite useful for setting sales, production for the aim of allocating resources to particular items (De Grauwe, 2010).
For example: if a company has created Samsung mobile app, then you should look that how many customers have acquired that app. If there are 20M active Samsung users and out of them half purchase at least a app per month, then you can extend it. You can also make an estimate that out of 10M active Samsung users who buy apps, 1% of them will buy your app. That would give you 100K new customers.
It is a method for presenting the likely futures for a company which is facing a lot of crucial problems by having an attempt to gain the range of probabilities that encourage the decision makers to depict the changes they would ignore else wise (Amer, et al., 2013).
For example: Scenario planning can be used by a farmer to know whether their harvest will be productive or not depending upon the weather. It will help them to forecast their sales as well s their future investments.
It is a method of forecasting the demand in the short term. It is suited well for the short and long term sales forecasts for the organizations. The results which are obtained by this method are quite accurate and realistic in nature.
In this method the companies or business organizations conducts surveys with their customers in order to identify the demands for their products and services which already exist and identify the likely future demand accordingly (Sun, B., & Morwitz, 2010).
The advantage of using this method is that it is very reliable in obtaining the relevant information, it is very effective in forecasting the short-run sales, and is also very effective when the customers clearly state what they expect from the business in context of their honest intentions, etc. The disadvantage of this method is that it is not an economical method.
For example:
| Survey Response | Actual Probability of Use |
| Definitely Will | 70 % |
| Probably Will | 35 % |
| Probably Will Not | 10 % |
| Definitely Will Not | 0 % |
| Survey respondents’ actual likelihood to purchase or try a product based on survey responses. | |
Take a survey for an entirely new product of 180 respondents. Out of them, 20 said that they will buy that product, 10 say that they will probably buy it, 25 say that they will probably not buy it and 50 says that they surely not buy it. From the above survey results and using the probabilities given above gives the following results:
(20 x 0.7) + (10 x 0.35) + (25 x 0.1) + (50 x 0.0) = 20
So it can be concluded that 20 out of 180 respondents will buy the new product.
There are basically four steps in this method of forecasting. The steps include selection of model, the fitting of model and the model valuation. The difference in the basic steps is identified afterwards.
Further it is seen often that one variation on model building sequence comes up when one additional data is required to fit a new hypothesized model which is based on model fit to the initial data. Data collection and experimental design should be put in between model fitting and selection of model (Khashei, M., & Bijari, 2011).
Amer, M., Daim, T. U., & Jetter, A. (2013). A review of scenario planning. Futures, 46, 23-40.
De Grauwe, P. (2010). Top-down versus bottom-up macroeconomics. CESifo Economic Studies, 56(4), 465-497.
Hasson, F., & Keeney, S. (2011). Enhancing rigour in the Delphi technique research. Technological Forecasting and Social Change, 78(9), 1695-1704.
Khashei, M., & Bijari, M. (2011). A novel hybridization of artificial neural networks and ARIMA models for time series forecasting. Applied Soft Computing, 11(2), 2664-2675.
Rowe, G., & Wright, G. (2011). The Delphi technique: Past, present, and future prospects—Introduction to the special issue. Technological Forecasting and Social Change, 78(9), 1487-1490.
Sun, B., & Morwitz, V. G. (2010). Stated intentions and purchase behavior: A unified model. International Journal of Research in Marketing, 27(4), 356-366.
Yeh, C. Y., Huang, C. W., & Lee, S. J. (2011). A multiple-kernel support vector regression approach for stock market price forecasting. Expert Systems with Applications, 38(3), 2177-2186.
