THE PROBLEM

Japan’s leading transportation company wanted to reduce significant costs on transportation fee with accurate demand forecast.

The client wanted to retain only those stores where viable future demand existed.

Customer’s existing method of demand forecast was manual, unable to yield consistent accuracy and had fluctuations due to external factors.

THE SOLUTION

Using the multi-algorithmic time-series forecasting capabilities of Findability.AI, the following indicators were forecasted:

• 45-day forward forecast for the numbers of packages to be transported between 18 pairs of locations.
• The total count of packages to be transported across Japan.

External factors such as macroeconomic conditions, seasonality and holiday flags were included in the analysis for improving the accuracy of forecasts.

THE SUCCESS

Forecast accuracy has been consistently maintained at above 95% for both, point-to-point and country wide predictions.

Accurate daily prediction of the number of packages to be transported, has helped the client allocate fleet effectively to optimize Vehicles and stores, thereby increasing customer satisfaction and a 7-digit savings in expenses in the current year.