Trying to understand and know what happens in the future has always been a topic of interest among humans throughout history. Will it rain tomorrow? How will this year's crop season be? How much of my production will I sell? All of us have desired to know what will happen in the future at some point. However, forecasting, as a science, is a growing field, especially among business. Obviously, knowing the future can increase profits. What is all this and how does it work?
Forecasting is about predicting the future as accurately as possible using all relevant information, such as historical data or a future event you know might impact the forecast. Before the use of statistical models our ancestors made judgmental forecasting. Those forecasts were based on experience alone and were related to weather and/or crop. Today judgemental forecasting is used when there is complete lack of historical data or when a new product is being launched.
Many businesses use some kind of forecasting to help inform decisions and provide a guide to long-term planning. Most of them, however, actually do it quite poorly. Many different factors need to be considered when building a forecasting model. Do you need short-, medium- or long-term forecasting, or maybe all of them? It’s not simply extending the horizon when making the forecast. It’s about selecting the best forecasting method for each horizon. You also need to decide what to forecast. Every product in a group or the group aggregated, every outlet or all outlets combined in a specific region?
It is possible to use a powerful automatic selection algorithm blindly for every product and every outlet and hope for the best, but that’s not the best solution if the goal is to maximize profit. When maximizing profits you want a forecast that is as accurate as possible. Just as in machine learning, forecasting can benefit from combining different forecasting methods into one super-forecasting model.
Businesses can save a lot of money if they can accurately predict sales or number of customers, as an example. By accurate sales forecasting you can avoid stocking too much, which can be quite costly, and you can avoid losing customers by not having enough in stock. The same holds true for a number of customers, since accurately predicting the number of customers you can better staff the store and thus either save money by not having too much staff or avoid losing customers by not having enough staff.
Forecasting can help different industries and businesses in different ways. What works for one company might not work for another. As forecasting techniques and technology become more sophisticated, forecasting becomes more accurate. This has created an enormous opportunity for businesses in terms of planning and execution with significant impact on the bottom line.
In a 2019 survey 67% of business leaders claimed that they expect to adopt advanced forecasting methods and they believe that it will give them competitive advantage. The same survey had 84% of those who had already implemented forecasting emphasise that they had increased profit as a direct result of advanced forecasting.
It is no overstatement to say that companies who will be late adopters of forecasting will lag behind their competitors and can have serious problems catching up with early adopters.