This global science company has a diverse product catalog of over a million items, marketed and sold through seven channels around the world. Product digitization was happening—one SKU at a time—but organized by the division that manufactured the product. As a result, e-commerce systems and trading partner syndication suffered from a lack of consistency and economies of scale, and customers had difficulty finding and comparing products.
Earley Information Science (EIS) worked with the client to organize products with similar attributes together, regardless of their origin, into a new customer-facing eCatalog. Using a machine-learning approach, we suggested clusters of “like” products into SKU groups, with each group sharing common attributes from the base product level and subgroups having unique variant attributes. Our process delivered digitization efficiency (even across multiple languages), and publishing control across brand sites, direct e-commerce, and trading partner syndication channels. All the information is now organized according to the way in which customers want to discover, compare, and buy.
Product on-boarding is now streamlined and more accurate, with global content reuse and faster time-to-market. The ability to find, manage, and sell similar products across the eCatalog assortment, regardless of where they originated world-wide, has been improved. The site now provides a customer-focused digital experience. Merchandising by product, industry, and application, consistent content for brand sites, e-commerce and syndication have been enabled. The move from a division-based organizational approach to a product-based taxonomy hierarchies and attributes creates a more uniform experience for the user. Together these capabilities offer the customers of this science company consistent presentation of its products and an efficient way to locate the ones they need.