Forecasting Gift Card Sales
A leading food service chain wanted to improve on a manual forecasting process to project monthly sales from retail and mobile gift cards.
The goal was to improve annual budgeting and to better-understand quarterly budget variance. It also wanted to better understand redemption.
Developed and deployed efficient and accurate time-series models to streamline and improve forecasts of future gift card sales related to two sales sources, one of which was related to the client’s loyalty app.
Augmented data sources with external data.
Assessed time-series forecasting algorithms and built a robust cross-validation framework to assess the forecast’s accuracy across multiple time periods and forecast horizons.
•A fine-tuned version of the Prophet algorithm incorporating seasonality, holiday effects, and changing trends over time
•A forecast process that used to take several hours each month can now be completed in less than a minute
•More confident cash flow predictions