Domain: Supply Chain Management
Service Area: AI/ML
Context
Demand Forecasting defined as the process by which the historical sales data are used to develop an estimate of the expected forecast of customer demand. Demand Forecasting provides an estimate of the of goods and services that customers will purchase in the foreseeable future. Client had the need to improve the accuracy of the forecast which initially was causing unavailability and consequently lost sales opportunities.
Solutions
Inventfor used Amazon Forecast to forecast and predict inventory needs at country level and provide the allocation of stock needed at item level for each location on a daily basis. We found that even with a very limited set of historical sales data, Amazon Forecast performed very well at forecasting for next days and weeks with no manual effort.
With Amazon Forecast https://aws.amazon.com/forecast/, sales prediction accuracy has improved by 50% compared to the existing forecasting solution. This means the client can stock the right products and save on purchase and stock management costs. Improved product availability also led to enhanced customer experience in the long run and minimize miss on sales opportunities.
Outcomes
Amazon Forecasting had achieved in a short time:
- 50% more accurate forecasts with machine learning.
- Reduce forecasting time from days to minutes
- Eliminate manual work
- 90% cheaper solution