Abstract:
Food waste is a global problem that undermines industry profitability, contributes to climate change and, when considered in conjunction with food poverty, is a potent symbol of the inequalities and inefficiencies found in contemporary food systems. With accurate data on food retail waste in New Zealand absent, this research is motivated by sales data from a New Zealand grocery retailer. Times series modelling methods were used to examine how weather data can assist with the forecasting of perishable food sales. The process identifies a number of sales series with statistically significant weather correlations. Validation came from a small number of these relationships found to reduce forecast error measurements. Data limitations prevent definitive statements, though it appears use of weather forecast data can improve the demand forecasting process and reduce losses incurred in perishable food retail to deliver both increased profits and societal benefits. However, data availability suggests refining data collection, management, and operational processes will bring greater inventory efficiency gains. The link between discounting, weather, and promotions is identified as an opportunity for interdepartmental collaboration as innovation will be required to solve the challenges facing perishable food retail.
