Case Study: Leading Distributor Company improves product search and categorization with Tredence's machine learning solution

Developed a Machine Learning Based Solution to Categorize Products and Improve Search on the Website for a Leading Distributor

Leading Distributor Company worked with Tredence to solve a major product categorization challenge across its inventory of 5M+ SKUs. With only 10% of items mapped to a product hierarchy, the company relied on a manual process for the remaining catalog, which hurt search relevance on the website, reduced customer experience, and contributed to revenue loss.

Tredence built a machine learning-based product classification solution that used product descriptions, vendor data, text mining, data cleansing, confidence thresholds, and business rules to automate categorization and prioritize low-confidence cases for review. The result was 30X faster categorization, 99% accuracy for high-confidence predictions, 95% overall accuracy, and an estimated $250K in annual cost reduction, while also improving site search and enabling incremental revenue.