Um... forecasting is a human's job? You have tools for that, don't you? Yes, but they are only effective if we (as humans) think about this process first. Moreover, forecasting is only a part of the complete S&OP (Sales & Operations Planning). Maybe you use dummy sales orders as forecasts in a system, but what is the use of this without consulting your colleagues in production and purchasing? Marketing campaigns, seasonal products, inflation; they all have an impact.
Being ad hoc does not lead to efficiency. Current raw material prices are no joke and without forecasting, you are more likely to run a surplus or shortage of raw materials. Or there is no production capacity, so customers don't get their orders on time. Forecasting what you will sell in the future is extremely important.
There are all kinds of tools that can help you do this. The starting point: making predictions based on historical data, preferably of three years or more. Based on that, algorithms recognise patterns and push them into the future. However, data from the corona era can be thrown in the bin; it is of little use. So how to make forecasts? With scenarios or AI! This is why we are investing in 'Garvis.ai': the world's first Bionic Demand Planning system that thinks along with you on the basis of AI.
S&OP can be challenging. However, don't let that stop you from starting it, and developing it further. Gartner has developed a model to determine the maturity of your S&OP capability. The most important first step: someone has to take responsibility for sales and operations. After that, you can deploy tooling, and ask people to share their knowledge with the system. You also need employees who can interpret the data. Only then will you be able to take the right actions for purchasing and production. So forecasting remains a human task.
Maarten Cordenier
Partner, Delaware
Source: Vakblad Voedingsindustrie 2023