Reducing food waste and improving planning processes: time series analysis for a restaurant for those in need

The "Mannheimer Platte" is a restaurant that offers affordable lunches to everyone, especially those in need. The project with CorrelAid aimed to optimize and evaluate internal processes within the organization. The Mannheimer Platte team needed insights into the factors influencing meal sales and demand to improve procurement planning. The data consisted of 17 years of daily time-series records, including variables such as the number of meals and drinks served and total revenue. The data was stored in Excel files. The CorrelAid team assisted with data analysis using R and SAS, cleaning and preparing the data. The project provided insights that confirmed volunteers' intuition about meal popularity on specific weekdays and revealed patterns, such as reduced orders and increased use of vouchers toward the end of the month. The organization learned the value of data and its potential for decision-making. The project results were shared through a handover workshop and presentations, including easy-to-understand heuristics derived from statistical models.