In our daily routine, we try to forecast lots of issues from weather to inflation, from match results to sales evolution, from life expectancy to agriculture outputs. A good forecast helps you to step forward, to act timely, to utilize opportunities, save more and earn more.
However, sometimes a monkey can make more credible forecasts with a dartboard than a human. But we don’t ‘sometimes’ need accurate forecasts, we need sustainable accurate forecasts based on facts and figures.
Each of us has an interpretation of issues based on our former experiences, based on our culture, and based on many other external and internal influences. Sometimes we are optimistic, and sometimes pessimistic. However, we need realistic forecasts at all times.
A good forecast makes a great difference: a good weather forecast helps farmers to harvest more products, a good stock price forecast helps investors to have a higher return on their equity, a good sales forecast helps companies to have better production planning. A good forecast helps improve profitability!
In this article, our focus will be only on sales forecast.
The first question to answer is ‘why do we need a good sales forecast?’.
Mainly due to three reasons:
- Tactical and operational in the short term (0-3 months) for an effective supply chain management
- Structural Forecast: Extending targets: annual budget process
- Strategic Forecast (> 3 years): Planning investments, investing in future growth, capacity/demand planning
A hands-on and visionary decision-maker (sales, C-level executive) knows the dynamics of a business and can forecast future product demand at the segment level. However, if hundreds and thousands of SKUs are involved, being a visionary and hands-on decision-maker will not help you to make an accurate forecast at all levels of SKUs and hierarchies due to the sheer complexity of the problem.
A good sales forecast at levels of a company’s product hierarchy will help you to optimize your inventory, manage the operating/working capital better, identify outperforming SKUs, manage product prices better and achieve higher levels of profitability.
In order to do reach these goals, you need a user-friendly tool that combines state-of-the-art statistical, machine learning, and artificial intelligence methods. The current business environment requires accuracy at all levels, not just at the level of a product segment and based on biased personal opinions.
AI and machine learning backed with statistical/econometrical models is the best solution in today’s world in generating a good forecast from big data.
To finish with 2 famous quotations from John Kenneth Galbraith: “There are two kinds of forecasters: those who don’t know, and those who don’t know they don’t know.” And from Ruchir Sharma: “The old rule of forecasting was to make as many forecasts as possible and publicize the ones you got right. The new rule is to forecast so far in the future, no one will know you got it wrong.”
Today AI and machine learning are there as a factual rule which is the third kind of forecaster that is the technology. This recent rule is to forecast accurately for one time and in a period short enough to justify its results.