More than data: it's networked BI that enables a "single version of the truth"
Data governance is a broad topic that people seem to agree is important, but still argue about the details. Specifically, it’s hard to find an IT professional or BI and analytics user who doesn’t believe that good execution on data governance will yield the “single version of the truth” required to make confident, accurate business decisions.
But what exactly is a “single version of the truth?” That’s the point in the conversation where the metaphorical gloves come off.
Data semantics are everything
Many IT and data management professionals posit that a single central data repository, such as a large cloud data warehouse or a data lake, is the definitive “single version of the truth.” This is easier said than done, as most enterprises are still rife with analytic silos created over the last decade by data discovery tools, as well as stubborn legacy BI silos.
In either case – the nirvana of a single cloud data warehouse, or the reality of multiple analytic data silos – there’s still another hurdle between enterprise data and confident, accurate business decisions. It’s the semantic data model, a conceptual data model that includes semantic information that adds a basic meaning to the data and the relationships that lie between them.
Birst’s semantic layer defines “a single version of the truth” with a common, reusable set of data definitions for key business terms such as:
- Cost of goods sold
Without a reusable set of data definitions, which can be used and re-used by everyone in your organization, BI users will make their own. For example, the definition of “profit” can be quite complex. Does it include returns or does it not? Is profit calculated before or after some other type of financial processing? It’s easy for business users to overlook these accounting nuances and, thus, base their decisions on improperly calculated numbers – or spend time at meetings arguing about whose definition is correct. Birst enables those data definitions to be defined once and used on a repeatable, consistent basis by everyone in your organization.
Connecting definitions to underlying data
Clearly, the underlying data that rolls up into the data definitions and, in turn, the business terms must be accurate. In most organizations, that data comes from multiple systems that are extremely difficult, if not impossible, to maintain consistency between. A BI and analytics platform such as Birst will reconcile inter-system differences and make sure the data, and the business terms presented to the user, are always consistent. For example, if two systems have different definitions of “profit,” Birst can reconcile the differences and map the correct data to the user.
Data definitions and business terms can also be standardized at a departmental level. As long as the process is governed and managed, organizations will have made a conscious decision to create a “single version of the truth.”
You can learn more about Birst’s unique networked BI approach to data governance in the white paper Governance in the age of data discovery. Read it to learn how to roll out common, reusable definitions across the enterprise in an agile way.