Logistika.cz – Coronavirus 'smashed' predictive models


Proper stocking is crucial for traders. Previously, retailers had to rely heavily on their own judgement and a great deal of manual work to plan. Today, sophisticated systems take into account all historical sales and model future trends with the help of additional data inputs, which helps them to predict demand. However, the coronavirus pandemic has "disrupted" demand causing the ordering systems not adequately responding.

Demand for some goods has suddenly dropped, while for others it has risen sharply, making it impossible to predict sales even for goods for which they were previously reasonably accurate. Systems are therefore unable to calibrate in this situation without human intervention. The problem is exacerbated by the current delays in the delivery of goods from overseas.

"The inability to correctly predict demand causes huge problems and losses for retailers, both in terms of lost revenue due to unavailability of goods and, conversely, expensive disposal of unsold items," says Roland Džogan, CEO of start-up Ydistri and an expert in replenishment systems to ensure continuity of supply.