A global food company contracts locally sourced potatoes before the growing season to support the supply of 70-85% of their raw ingredients for one of their potato crisp factories.
To mitigate risk in seasons of lower than expected yields, the company also procures potatoes on the open market from other countries in Europe.
Potato yields may fluctuate annually by more than 20%, while open market prices can swing by 200% over the course of the year. Because of this, the company was interested in earlier and more accurate yield estimates for their locally sourced potatoes to make better decisions for their open market purchases.
The company became a client of Trellis, and their potato crisp factory joined the Trellis platform at the beginning of the 2018 growing season.
Over the course of the entire season, the client’s procurement and agronomist teams observed Trellis’ AI-powered, real-time yield predictions for each of the contracted suppliers from whom they source potatoes.
The potato chip company observed in the Trellis platform already in month three of the growing season that their yields at harvest were predicted to be 15% lower than the production line had planned to receive.
The procurement team immediately turned to the open market, and sourced the required supply of potatoes. These potatoes were supplied to their production line in a matter of weeks, and the company avoided a costly inventory shortage.
Shortly after the client procured their missing supply on the open market, the price of potatoes increased by 40% (from 220 to 300 €/t) due to an extreme shortage in the European market. Had the client not leveraged the Trellis prediction to procure earlier, their monthly costs of goods sold would have increased by 8%.