top of page

Case Studies

Distributor Case Study

Project Summary

The client was an industrial distributor who recently purchased two regional competitors and needed to merge the product catalogs. The project began with a full inventory and summary analysis of existing product data repositories. Each repository was then loaded to SkuDB, where data sets were integrated into a new master taxonomy, including a two-tier data model for Products and SKUs.  

​

SkuDB’s AI-enabled tools then classified items across the newly created categories and hierarchy. Once completed the new design was finalized with subject matter experts (SMEs).  Global attributes and code sets were created and launched in SkuDB, and navigation attributes were assigned to categories based on existing data including revenue and competitive research.  Product attributes and schemas were created and normalized in SkuDB according to style guidelines approved by Client stakeholders and enhanced by SkuDB's intellegience layer.  Taxonomy mapping rules were auto-derived from item classifications and built into onboarding workflows. Data delivery files were exported from SkuDB in load-ready format for Client's PIM.​

Results

•    Master taxonomy category count reduced by 30% 
•    Ecommerce taxonomy average depth reduced by 40% 
•    Misclassification rate reduced from 2.5% to 0.2% 
•    Navigation attribute count increased by 50% 
•    Navigation attribute fill rate increased by 20% 
•    Search relevancy increased by 25% 
•    Item onboarding time reduced by 20%

A Closer Look 

  • Two-tier data model (Product-SKU) with images, descriptions, and attributes  assigned at both levels 

  • Enhanced PIM attribute metadata designed to enable  left navigation, search indexing, data governance, and  other business requirements 

  • Revised unit of measure attributes to clarify ecommerce package and order quantities 

  • Search thesaurus and search redirects to improve keyword-level search relevancy 

  • SKU-to-SKU and Product-to-Product relationships defined and normalized for ecommerce display 

​

Manufacturer Case Study

Project Summary

Manufacturer was a large international industrial conglomerate with a variety of disconnected legacy PIMs.  The project team initially built proposed level 1 (L1) categories in SkuDB for the top-level business segments in an enterprise-wide master hierarchy.  Because of the easy to understand interface, the structures were presented in the application to project leaders and executive sponsors.  After receiving approval for the L1 structure, the project team embarked on a 2-week socialization and normalization campaign with brand and marketing teams, using SkuDB as its main presentation platform, and documented suggestions and issues in real time during review sessions with stakeholders.  

​

After completing 3 pilot categories end-to-end, the project team then used SkuDB to design a full taxonomy and work with subject matter experts (SMEs) in product engineering and marketing to refine and approve the design, which included an ecommerce hierarchy with navigation filters.  Hierarchy mapping rules and a search thesaurus were then created based on exports from SkuDB and additional client inputs. Hierarchies, schema, mappings, and data quality rules were loaded to the client's implementation platform.  In parallel, an updated master hierarchy and updated product descriptions were loaded to ERP.

Results

  • New global structure for master product data and ecommerce display hierarchies

  • Proof of concept launched for ecommerce on time and on budget

  • Reduced onboarding time by ~25%

  • Search relevancy increased by ~20% 

  • Attribute duplication eliminated in electrical products segment

  • Ecommerce taxonomy average depth reduced by 30% 

  • Reduction in maintained websites by 50%

A Closer Look 

•    Flexible depth ecommerce hierarchy with definitions, help text and style guidelines created
•    Two hierarchy (master-display) model for clean data governance 
•    Navigation attributes including critical metadata for ecommerce generated
•    Search indexing, search thesaurus and search redirects enhanced for improved UX
 

Retailer Case Study

Project Summary

Client was a top 10 ecommerce retailer in home improvement market and was launching a new ecommerce portal in the next 6 months. The master taxonomy was built first in SkuDB, and SKUs were classified leveraging both AI-driven workflows and working sessions with subject matter experts.  In parallel, the design and construction of a display taxonomy commenced in SkuDB.  After hierarchies and product classification were approved by client stakeholders, taxonomists began navigation attribute schema design in SkuDB, including competitive analysis, attribute definitions, controlled value lists, sequencing and metadata authoring. 

​

After navigation attributes were approved, existing data was migrated into the new design, including legacy data, manufacturer data, GS1 data, and data gathered from image assets.  After legacy data migration, an offshore team began data sourcing from manufacturers in order to fill data gaps and improve fill rates.  The sourcing team leveraged attribute definitions, controlled value lists and supplemental training documentation, and a double entry process was used to increase data accuracy.  Final data was normalized by taxonomists and then uploaded to a homegrown PIM.

​

After data implementation, content from SkuDB was leveraged to stand up a new product data governance program, including style guidelines, a preferred term dictionary, attribute and category definitions, controlled value lists and an issue log.

Results

  • New ecommerce site launched on-time and budget 

  • Sales through ecommerce portal increased $100 million in first 12 months 

  • Ecommerce capabilities extended to 225 new stores 

  • Average NPI time for ecommerce reduced by 2 days (~30%)

  • Misclassification rate reduced from 3% to 0.25%

  • New data governance council established 

A Closer Look

•    Two hierarchy (master-display) model for clean data governance 
•    Single-tier item model to allow for quick NPI and simplified management across brands 
•    AI-assisted item classification to speed onboarding
•    Navigation attributes designed with critical metadata for ecommerce 
•    Product data formatted into PIM load template 

bottom of page