DATA SERVICES HELPS A COMPANY SHARE INCENTIVES

THE CHALLENGE

A global technology corporation was looking for assistance with two upcoming data-based projects.  They requested:   

• Oversight of data services within their organization    
• Maintenance of a database containing records from computer retail stores

THE SOLUTION

Cayuse provided data gathering, cleaning, updating, and mining services for the Client’s Worldwide Channel Partner Program. The Client’s partners either buy directly from them or purchase through large distributors. The partners who buy indirectly account for about 44% of the overall global sales. A specialized incentive program offers those partners:  

• Partner training  
• Access to special support vehicles and conferences  
• Access to sample hardware  
• Funds for marketing  
• Partner rebates  

With support occurring in over 100 countries and in 30 languages, the challenge was keeping the partner demographic and business specializations clean, updated, and effective, with respect to country codes and location data.   

To keep the Client’s database up-to-date, Cayuse gathered, cleaned, and de-duped the data along with mining it for various requests. This information stemmed from a variety of sources including voice contact, email, web metadata, spreadsheets and database dumps. Various location data, including country postal codes and landmark locations were normalized, and made to be as intricate as discerning between abbreviations and full words, such as ‘Street’ vs. ‘St.’.   Cayuse engineers put integrity and matching efficacy into place by verifying and pairing the data with internal information for each partner.    

The tools used in this project include:  

• Data Stored in MS SQL (Microsoft SQL Database Server)  
• ETL (Edit/Transform/Load) performed using MS SSIS (Microsoft SQL Server Integration Services)  
• Normalization tables during the ETL process to verify that all data was consistent and uniform from refresh-to-refresh cycle.  
• De-dup logic performed using SSIS Fuzzy Logic Component with matches at 80% or higher auto-approved, versus 79% or lower verified by hand.  

THE OUTCOME

The strategic decision-making capacity of the partner database requires that the data is continually verified and de-duped. Partners are contacted biannually to verify their information. Before adding to the database, a test system is used for final checks. Every six months, the data which includes demographics, business specialization, sociographic, and geospatial information is refreshed. The database is critical to the financial success of the Client and is confidential, proprietary, and heavily guarded.