Customer demand behaviour depends on a variety of external factors. Accordingly, we developed a powerful daily forecasting model for customer demand and revenue that incorporates factors such as trends, holidays, seasons and days of the week.

Our observation: In the course of this project work and in others related to forecasting concerns, we often notice that our clients establish reporting on a monthly basis in comparison with respective months from previous years—not taking into account the impact of holidays. On one hand, this can lead to seemingly spurious effects; on the other, it can conceal significant changes that actually do exist. Regardless, the significance of previous year comparisons is reduced if, for example, the fact that Easter occurs in April during the current year but in March the year prior is not taken into account.

The model we developed considers the timing of religious and public holidays and can be leveraged not only for forecasting but also to adjust comparative value reporting at the monthly level.