Leading Auto Parts Retailer Inventory Replenishment Automation

Background

A leading automotive aftermarket parts provider with over 5,000 stores spread across North America serves both professional installers and do-it-yourself customer was in a quest to serve their customers better by improving product availability. The following factors motivated this company to rethink their stock replenishment strategy:

• Market demand for 200K+ parts
• Average product sells less than 1 unit per year
• There is high cost of carrying inventory
• Lack of availability in less than 2 hours leads to lost sales
• Recent merger has led to duplicate systems and associated challenges for data capture and processing

With this in mind, this company initiated a program to implement Automated Stock Replenishment that leverages integrated demand forecasting, proportional inventory management, and automatic order generation to create automatic replenishment plans and orders.

Challenge

Data preparation is an important and critical step before advanced algorithms for Automatic Stock Replenishment are applied to realize demand forecasting and order generation. Data from multiple operational systems (mentioned below) must be ingested, cleansed and transformed into a standardized data model. The result is consumed by AI algorithms to produce actionable insights.

• Point of Sale Data
• SKU Database
• Product Hierarchy
• CRM Data
• Store Data
• Inventory Data
• Weather Data
• Demographics
• Social Media Data

The challenging part of this problem was to design and implement a solution that could scale horizontally to handle the volume, variety and velocity of data from these systems.

Solution

pSolv designed and implemented a comprehensive solution including a big data technology enabled data processing platform with these key capabilities:

• Demand prediction: predicting part demand at any given store
• Assortment: optimizing part assignment to stores
• Replenishment: replenishment across supply chain
• Visualizations: to support business decision making

Outcome

• pSolv rapidly implemented a comprehensive end-to-end solution
• Performance of optimization process has been significantly improved – was taking 20 hours for one market now takes 25 minutes for all markets
• The solution provides end-to-end Supply chain visibility for power users