Restaurant demand forecasting involves analysing historical information using quantitative and qualitative methods, an exercise that is often inaccurate, highly dependent on staff’s experience and very time consuming. While digital demand forecasting solutions that automate and improve accuracy of this tedious process now exist, they offer a fragmented experience and in most cases are very expensive and complex.
Based on advanced machine learning algorithms and key influencing factors such as historical data, weather, holidays, vacation time and significant events, Prognolite predicts a restaurant future demand with high precision (92%), contributing to 1) reduce overproduction by 70%, 2) reduce total food waste by 50% (30gr savings per meal), and 3) adapt the staff planning to the real demand by suggesting the most appropriate work shift.
A feasibility study is needed to confirm that there is a real market opportunity for the product whilst the conditions for the commercialisation are beneficial.