Demand Forecasting
Knowing the future demand for your products enables you to act correctly.
What is demand forecasting and why is it important?
Imagine knowing exactly how many of your products you will need in the future. With this knowledge, you could take all the right decisions today: allocating the correct number of items in every store, everything being available so that your customers get what they need. In contrast if you have no accurate forecast available then multiple problems arise products where you underestimate the demand become out of stock, you lose revenue, and your customers move (dissatisfied) to your competitors.
Who needs demand forecasting?
Demand forecasting is mainly used in retail but all sectors in the product lifecycle chain need to forecast demand accurately:
- retail: to have the correct amount of each product available in store
- e-commerce: to be able to cope with increased online demand keep product availability high
- wholesale,
- logistics: to have the correct amount of each product available in store
- manufacturing also have interest in accurate predictions.
Also, usually several departments and roles are involved in forecasting:
- store management,
- supply chain managers,
- product developers and
- category managers.
Why is demand forecasting difficult?
Demand forecasting is complex. It has to be done for different time periods, at various levels. All forecasts must be harmonized to produce consistent numbers which can be shared across departments. In addition, business rules must be respected, and forecasts need to be integrated with ERP systems.
At the product or sku level demand is often intermittent and volatile, this makes forecasting based on expert knowledge or purely past data inaccurate.
Shocks and pandemics affect global supply chains. Since the outbreak of the covid-19 pandemic many sectors have experienced either highly increased demand or stood still during lockdowns.
Historical data show sales and not always demand. In out-of-stock situations the true demand is partially unseen.
