Footfall forecasting predicts how many people will visit a store, café or venue in a future period. Forecasts can cover the next hour, day, week or season.
The starting point is history: weeks or years of visitor counts by hour. Forecasts then adjust for patterns such as day of week, school holidays and public holidays, and for known events like sales, local festivals, weather and marketing activity. Simple methods use averages for the same period last year, while machine-learning tools weigh many factors.
Operators use forecasts to build rosters ahead of time, order stock and fresh ingredients, plan deliveries and set sales targets. Linking the forecast with POS data lets a manager predict sales and labour needs, not only visitors.
Accuracy depends on clean, consistent counts, so counting errors, sensor outages or changes in shopping-centre opening hours will weaken results. New stores and unusual events have little history to learn from. Compare forecasts against actuals each week and adjust. Recording unusual events, such as a road closure or a sensor outage, helps explain misses and improves the next forecast. A forecast is a planning aid, not a guarantee.
Better predictions reduce both overstaffing in quiet periods and shortages at peaks, and cut food and stock waste.