How can I forecast the next year accurately and economically?

PC:Google Image

Ameya and Anagha got married last year. They selected each other. Parents from both sides didn’t make much noise. Both of them are a ‘made for each other’ pair.

They are engineers and work with different organizations. Ameya is in planning and Anagha is in design.

Yesterday Anagha noticed that Ameya was not at ease. Something was wrong. When they sat together on the balcony Ameya broke the silence by saying that his annual plan was not approved. Anagha asked the reason. He said he prepared it based on last year’s achievements of different products. Anagha asked her whether he got the forecast from the marketing department. Ameya got angry. As an assistant planning manager, it was his privilege and responsibility to prepare a plan for the coming year. He never allowed anyone from any different department to interfere in his work. Anagha knew that he was a self-centered person.

Anagha wanted to explain it to him but was reluctant to hurt his ego.

The doorbell rang. Oh, it was Aniket, Ameya’s elder brother. Aniket works in a corporate office of a reputed company. With his maturity, Aniket understood that there was something unpleasant in the air.

After a small gossip session, Aniket asked them how the day was. For Anagha, it was quite nice. She was working on a new concept with the help of AI.

Ameya was restless. He was fuming. He narrated the comments of his boss and colleagues on his next year’s annual plan. Based on his education during his MBA, he said, ‘Bhaiyya, forecasting is a method of foretelling a future event or future data by considering patterns and detecting the trends in past and current data. This requires mathematical approaches to lead to predictions. Am I correct?’

Aniket nodded and said, “Yes. You are correct, but it has to be supported by the marketing personnel. They are the ears and eyes of any organization. They move into the market. They meet past, present, and prospective customers. The market is never stationary. It is always better to consult them to understand the market trend. They provide us with the Sales and marketing forecasting.”

Aniket continued, “Let’s see the advantages of the basis of forecasting. Demand forecasting in business helps in strategy, location, layout, and capacity planning. This normally gets discussed at a higher level. Now, if an organization has two or more locations for manufacturing the products, the CEO will decide which area is suitable depending on transportation and ease of handling the products. If the demand is tremendously high, the company may be required to change the plan's layout. And you know how difficult it is to change the layout. The final plan comes to your planning department for further action. Then, you consider the past data, which provides trends used to forecast future trends. This helps to decide on products or services to be continued or discontinued. Past data helps in planning almost correctly.

Further, based on your forecast, you can estimate the fund requirement to ensure the availability of raw, semi-finished, and bought-out items required for the sub-assemblies and final assembly.You may have to arrange special and extra toolings and fixtures.

Accurate prediction of future demands of products and services increases the operating efficiency. Correct forecasts improve the utilization of the plant.”

Aniket asked Ameya whether he was aware of the types of forecasting. Ameya said, “Yes. I had studied in college. There are two types of forecasting: Qualitative and Quantitative. Under the quantitative types, there are two subdivisions: Time service and Casual.”

The discussions took a lot of time and Aniket remembered that his wife Ananya must be waiting for him to have dinner together. He was about to leave for his home but Anagha stopped him. She had already phoned her Ananya Bhabhi to join them for dinner at their place.

So Aniket continued, “A few important characteristics of quality forecasting data consist of i) the forecast is normally concluded on personal judgment or some external qualitative data ii) the forecast is subjective. It solely depends on the experts whether they are optimistic or pessimistic. ii) The plus point of this method is that it is speedy. iv) These methods are normally used for individual products or product families, not entire markets.”

“You must be aware of methods of qualitative forecasting” Aniket inquired.

“Yes, yes Bhaiyya,” Ameya said, “They are i) Market surveys, ii) Delphi or Panel consensus, iii) Life cycle analogies, and iv) Informed judgment. But will you please explain to me Bhaiyya?”

Aniket started, “Market surveys are normally structured questionnaires sent to prospective customers. They expect their opinions regarding the product and also take the opportunity to know their requirements in advance. If analyzed properly, they can be quite practical. The onlything is that they are expensive and time-consuming.”

Aniket was now in a good mood to explain further. He said, “Delphi or Panel consensus uses panels of experts in the market for which the survey is undertaken. The main difference between the two methods is that panel forecasting assembles the experts in a meeting for discussions, whereas the Delphi method permits a series of forecasts by each expert.

There is another method called ‘Life cycle analogy forecasting.’ This method is used normally when a product or service is new. The questions here are different: what is the time frame? How rapid will the growth be, and how long will growth and maturity last? How large will the demand and how rapidly will it decline?

They also used ‘Informed Judgment.’ Though it is a common forecasting method, it doesn’t yield good results.”

“In quantitative forecasting,” Ameya said, “the main characteristics are i) the foundation of this method is the relationship between variables. It is assumed that one measurable variable is the source of change of another variable in a predictable manner. ii) There is another assumption of causality that the causal variable can be measured correctly. This measured variable which causes others to change, is often called the ‘leading indicator.’ If good indicators are developed, then this method will bring excellent forecasting results. iii) As an extra benefit of the process of developing this model, one who develops can lay his hand on additional market knowledge. iv) These methods are rarely used for products, but frequently used for the entire markets or an industry. The drawback of these methods is they are time-consuming and very expensive.”

Aniket wished to wrap up the discussions on the types of forecasting, and he said, “Some of the normally used methods of causal forecasting are: i) Input-output modes, ii) Economic models, iii) Simulation models, and iv) Regression”

“In Quantitative Forecasting-Time Series,” Ameya said, “the methods are based on past demands, Intrinsic Forecasts, Random Patterns, and Trend Patterns. And still popular method is- the Moving Average Method. There are again two variants- Weighted Moving Average, and Simple Exponential Smoothing.”

But Ameya, “Are you aware that there are two forecasting errors in business?” Aniket asked, “They are Mean Forecast Error (MFE) and Mean Absolute Deviation (MAD).”

Two ladies were restless. Anagha proposed that they all should have dinner first and then both brothers could continue the balanced discussions.

Aniket and Ameya agreed.

After the dinner, Aniket stated, “In an industry, cost against accuracy is a very important factor.In economic forecasting for business, it is seen that a higher need for accuracy attracts higher costs of developing forecasting models.  Therefore, it is usually asked for a budget of money and manpower to develop the forecasts. The top management is also inquisitive about possible benefits expected from accurate forecasts and the possible costs of inaccurate forecasts.

“Let us wind up with the fundamental characteristics of Forecasts and characteristics of a good forecasting system,” Aniket was getting late. He said, “ Please remember that i) Forecasts are normally wrong, ii) They are more accurate for groups or families of products, iii) they are accurate for shorter periods, iv) every forecast should include an estimate of error and v) forecasts cannot substitute calculated demand.”

“The characteristics of a ‘good’ forecast system are i) accuracy, ii) low cost, iii) online capabilities, and iv) facility to connect into the present database management system.”

There were excellent discussions on ‘Forecasting’ between the two brothers. Anagha and Ameyathanked Aniket Bhaiyya and Anjali Bhabhi.