Intel IT’s master data strategy goes beyond selecting a master data management (MDM) solution to include key elements that will drive business efficacy.
Imagine a customer list being maintained by two business units, each unaware of the other’s updates. Or what if Finance and Manufacturing have differing views of supplier data? This siloed approach to data create problems, such as a customer falling through the cracks, leaving a potential deal on the table. Or the wrong product could be shipped to a factory, delaying product delivery.
Intel IT is committed to creating a single version of truth that is available to all business units and employees. Our master data strategy is designed to provide the highest quality master data to the enterprise in a way that will flex and scale as the company evolves. All the pieces of this strategy mesh to make high-quality master data sustainable for the long term:
- Establishing Data Ownership
- Defining a Master Data Policy
- Creating a Master Data Glossary
- Defining a logical data model (LDM)
- Selecting an MDM
- Integrating an MDM with Third-Party Data
- Building a Trusted record of reference (ROR)
- Deploying Reusable APIs
- Governing Adoption of the Master Data Strategy
Terms to know:
- Record of origin (ROO). For master data, this is typically an MDM solution. This is where the data gets created with as much accuracy as possible.
- Record of reference (ROR). A data warehouse, data hub or data lake that serves as a single source of truth for a master data domain.
We avoid having one ROR feeding another ROR—primarily because it introduces an extra data hop where mis-translation can easily occur. We always have the ROR obtain the master data directly from the ROO. This helps ensure that the same data is made available to all corners of the company while continuing to reap the benefits of data warehouses, like those with in-memory database capabilities.
Our ROOs and trusted RORs are built using Intel® processors, solid state drives and Ethernet products. We make master data available everywhere the company needs it by having trusted RORs both on-premises and in the cloud.
Properly mastered data is important to Intel’s success, which relies on data to fuel growth and transformation. Bad master data can disrupt reporting, analytics, dashboards and even machine learning model accuracy. We expect that our efforts to create clean, accurate master data across Intel will result in higher business efficiency and effectiveness.
There is more work to do, and we will undoubtedly revise our LDMs, modify APIs, get new data sources and hone our governance policies. But we are well on our way to creating reliable sources of master data and democratizing this corporate asset so every app, report and algorithm across Intel is getting its master data from authorized ROOs.