Every percentage point improvement in forecast accuracy is worth millions of dollars for a billion-dollar CPG manufacturer
Challenge
Low short term forecast accuracy at a location level causing inventory mismatches
Solution
A.I Demand Sensing utilising weather and global events data
Client
One of the world’s largest beverage manufacturers
Impact
9% forecast accuracy improvement
Before
The client had a strong top down forecast using SAP which was then disaggregated to produce a forecast value at the lower levels in the product/location hierarchy. Unfortunately, the forecast accuracy at the lower levels was unsatisfactory, resulting in a suboptimal inventory holding across their DC locations
Remi AI
Machine Learning
Demand Sensing
Weather Data
Events Data
Promotions Data
Hierarchical Forecasting
After
Forecast accuracy was increased by 9% on average. This was driven by a combination of machine learning forecasting at multiple levels in the product hierarchy, as well as the inclusion of events, promotions, and weather data.
.png){kind=logo}