During a recent Webinar entitled Data Quality: Getting Started the Right Way, David Wells, an industry consultant and former education director with Renton, Wash.-based The Data Warehousing Institute(TDWI) discussed the importance of data quality in an organization. Hesaid that in the 80s and 90s, data was not perceived to be owned byanyone in particular, rather it was an enterprise asset belonging tothe organization as a whole. But Wells said the problem there is everypiece of data needs an owner who will make decisions on the relativeimportance of its quality, privacy and security. Otherwise, it will be“orphan data” doomed to poor quality.
To avoid this issue, Wells suggested that roles and responsibilitiesbe created around data quality. One such role, custodianship of thedata, is typically assigned to IT to ensure the data is stored,backed-up and secure. Another role, ownership of data, is typicallyassigned to the business. And finally, there’s stewardship of data,typically assigned to both IT and the business, which he described asresponsibility without authority to build consensus and push people towork toward data quality.
The role of stewardship of data is also mentioned in recent research by Aberdeen Group Inc. entitled One Version of the Truth 2.0that discusses a positive correlation between “best-in-class”organizations and technology investments. Aberdeen describesstewardship as more than just ownership in that “data is managed,cleansed, stored, and delivered according to the policies, guidelinesand governance specifications as defined by a formal committee.”
Wells’ and Aberdeen’s definitions of stewardship have one thing incommon that stands out to me: a driving force or champion behind anorganization’s quest for data quality. While it’s necessary thatdifferent components of data quality be cared for by whicheverdepartments are best suited for the job (like IT is best suited toensuring storage and security), I agree that a force must be present tothread these components together thereby rendering the enterprise-widescope that it deserves.
But the Aberdeen research goes a little further to include a formalcommittee as part of stewardship to oversee the data quality processand define guidelines and governance specific to that organization. Acommittee approach certainly makes sense especially if the members arerepresentatives from different departments who can each contribute adifferent perspective on data quality. But while a formal committee isgreat for a large enterprise, I’m guessing it’s not an option for asmaller business. But that doesn’t mean they can’t have an equivalentresource in the form of a single individual, perhaps like a dataquality strategy officer.
