1. Rationalizing and sanitizing to achieve high-quality master data
To help our client manage their master data with ease, our teams first identified areas where they could make significant gains. Next, we analyzed 309 attributes of items across four websites and saw the scope of simplifying the data model by 42%. To achieve this, we eliminated duplicate data and consolidated the data model to 132 attributes in total.
Further, we analyzed 70,000+ items across distributed enterprise systems, which were initially classified into 320 categories. Then, we rationalized them to 172 item classes, which helped us establish a consolidated core taxonomy. We achieved by reinstating the parent-SKU inheritance relationship at the SKU level. In addition, we leveraged our RPrIm accelerator templates, which helped us implement mass configuration of setups, ensuring minimal disruption and rapid delivery for our client. Finally, 70k+ items were migrated from distributed on-prem systems to Oracle Product Hub Cloud.
2. Achieving real-time system interfacing with Oracle Integration Cloud
Because our client was still leveraging on-prem technologies like Product Information Management systems, we implemented Oracle Integration Cloud (OIC) to build real-time integration and reverse interface with the Product Hub Cloud. This entailed installing an OIC agent on the client's network to integrate OIC and on-prem databases. OIC implementation brought speed and responsiveness into the core business logic and enabled complete automation of business criteria and processes for various scenarios in a scalable fashion.
3. Closing the loop with a data stewardship function and proactive measures
Maintaining high-quality master data is a continuous process that calls for adequate process guardrails and people practices. To help the client prevent lapses and drift in master data quality, we helped them establish a dedicated and scalable Product Data Stewardship function and recommended optimal processes, defined standards, and policies.
We also defined product governance processes with 100+ business rules underpinning change management workflows. In addition, we implemented a reusable error handling mechanism in OIC to notify the right teams, thereby helping them proactively diagnose core systems in cases of failures. Lastly, we used a completeness score utility to generate a measure for item quality and implemented a criterion to prevent the fall in the quality score of an item below 70%.