Client
- Martinus Major East European omni-channel e-commerce bookstore managing multiple online web properties and operating 15+ brick and mortar stores. Martinus is a recognized retail category leader and has received awards including Deloitte Fast 500 EMEA and MasterCard Retailer of The Year.
Challenges
-
Inventory replenishment process and purchase order system relying on empirical rules and subjective judgment resulting in sub-optimal inventory levels and stock-outs
Large inventory spanning several hundreds of thousands of products making it tedious and difficult to correctly estimate product demand on the SKU level
Orders with fulfillment times ranging from days to weeks for many products
Solution
-
Automated machine learning based demand forecasts on the SKU level integrated in customer’s order management process
Seasonality modeling is performed with unsupervised machine learning, which further enables predictions for new products, as well as decreases variance of forecasts resulting from noise in individual SKU sale data
API fully integrated in matter of days, including historical and daily data synchronization and integrating demand predictions into customer’s internal systems
Results
-
Ratio of products expedited on the day of placed order increased by 84%
Improved average order fulfillment time by 14%
Reducing sub-optimal replenishment choices due to subjective manual sale forecast computation
Increased shopper satisfaction by guaranteeing shorter delivery times, thus improving brand perception and increasing customer loyalty
Customer quotes
-
"Integrating Algopine Sale Forecasting API brings benefit both to our product specialists by saving their time, and to our sale managers by providing them a cutting-edge tool for improving the complete inventory management process."
“I’ve noticed a dramatic spike in our same-day order fulfillment ratio just few days after fully integrating the Algopine API predictions in our processes. This improvement was so prominent, that for a moment I thought we had wrong numbers in our analytics dashboard. After confirming with our sale director, the spike was indeed directly trackable to our recent Algopine Sale Forecasting API integration.”