Implementing data lineage is a complex task including challenges such as managing multiple legacy systems, understanding data flows, responding to ongoing regulatory change, winning management buy-in and securing dedicated budget. Done well, however, it can provide significant benefits.
The challenges and opportunities of data lineage were discussed during a recent A-Team Group webinar sponsored by ASG and titled, The Art of Data Lineage. The webinar was hosted by A-Team editor Sarah Underwood, and joined by Jesse Canada, vice president of business metadata, rules, and reference data management at Citizens Bank; Sue Habas, vice president of strategic technologies at ASG; and Yetkin Ozkucur, global practice vice president for data intelligence at ASG.
An audience poll setting the scene for discussion found 29% of listeners planning data lineage implementation, 31% implementing in-house solutions, 14% implementing vendor solutions, 9% having already implemented solutions and 16% having not yet made any progress.
Moving on from here, Habas defined data lineage as the discovery of where data originates and how it moves from system to system, and noted the requirement for end-to-end lineage to validate data, as well as the need for data lineage to be tied to business needs so that necessary data can be found quickly, a particular concern in reporting for regulations such as BCBS 239 and the US Federal Reserve’s Comprehensive Capital and Analysis Review (CCAR).
Data lineage projects tend to major on regulatory compliance, but can also support other data management functions, such as deciding what data is important and eliminating redundant data. Canada described the need for top of the house sponsorship too secure buy-in for projects and noted benefits of meeting regulatory requirements and handing data ownership to lines of business. Another benefit is the ability to support modernisation and migration programmes. Considering objections to projects, the panel talked about the length and complexity of projects, and fear of the extent of information that must be understood to produce good data lineage.
Looking at key milestones to a successful outcome, Ozkucur described a three-step process covering: preparation, including defining data scope; the use of automated tools to capture data lineage, but understanding that tools may not produce 100% accuracy; and the need to maintain data lineage by working within a data governance framework to continually identify changes and notify consumers.
Finally, offering a little advice to listeners, the panel suggested data management practitioners should acknowledge the importance of metadata in data lineage projects, let the most important business data guide implementation, and view data lineage as an ongoing rather than one-off effort.
Listen to the webinar to find out more about:
- Challenges of data lineage
- Best practice approaches
- How to win buy-in
- Business use cases
- Milestones to success
