Many Asia Pacific companies use a responsibility, accountability, contribution, and informed (RACI) matrix for data governance. A concept published in 1956, its ability to assign who’s responsible for what area in governance drives its appeal.
For example, RACI charts allow companies to understand where and how master data decisions are made. It also keeps everyone abreast of any significant policy decisions, standards, rules, or essential items related to master data management.
The success of RACI in adding clarity to complex projects is seeing many companies extending the concept to data governance. You’ve data owners, data stewards, data producers, data consumers, data custodians, and a data quality team (supporting data quality across the company).
Some see RACI charts as outdated. Paula Martin, the chief executive officer and chief creative officer at the Matrix Management Institute, unceremoniously urged companies to flush their RACI charts down the toilet.
She pointed to the RACI suppositions as to the source of her concerns. It assumes that an individual will make all the decisions and a single person is accountable for each decision or goal. These assumptions come from Vertical Management 1.0 concept, where individuals drive an organization.
This creates three significant issues. First, it is not practical. Very few companies have the luxury to build data governance teams within DataOps. Many teams are still small (and facing relentless talent crunches). Appointing data owners, stewards and custodians is not a luxury for many lean and overworked data teams or business teams.
Having compliance take control over data governance can create friction and delays. Calls to expand data governance with data ethics are also creating fault lines with the established RACI matrix.
The second is that RACI is not Agile. And Agile is what drives many DataOps teams.
Some believe they can work together. Christopher LeCoent created “F” for the facilitator position. Mike Adler defines the facilitator as someone who facilitates communication and information across the team and aligns it with the SCRUM master role.
Either way, RACI is not designed with agile teams in mind; it needs to be retrofitted. Also, many DataOps teams seldom work linearly; they constantly send input and share concerns while experimenting with new ideas in huddles. So instead of assigning individuals, the entire team should be responsible.
Lastly, RACI won’t solve dysfunctional DataOps teams, wrote RACI Solutions. There is no shortcut to ensuring every team member understands their specific scope of work (even if it blurs in an agile team). It also won’t solve low morale, leading to responsibility seen as a liability and becoming a target for blame.
But RACI is time-tested, a key reason why many companies keep adhering to its principles. Project managers always highlight its ability to offer clarity in complex, cross-functional projects as a significant benefit. And many DataOps projects are becoming ever more complex.
So, how can RACI be made to work for agile DataOps teams? Or is there a better answer? While answers to both questions are not evident, there is an interim one: DataGovOps. And before you “face-palm” or groan over another Ops term, it is one that I think is long overdue.
The most significant advantage of DataGovOps lies with automation and the use of ML to drive governance decisions. For example, if someone exports sensitive data from the sandbox environment, the data owners, stewards, and custodians can be alerted. If there is any policy violation, it can trigger real-time alerts. This allows DataOps teams to focus on their primary job and data governance rules to be easily updated (or even customized for confidential projects).
No, DataGovOps is not the cure-all current RACI’s challenges. But it offers some breathing space as DataOps teams figure out what comes next after it.
Winston Thomas is the editor-in-chief of CDOTrends, HR&DigitalTrends and DataOpsTrends. He is always curious about all things digital, including new digital business models, the widening impact of AI/ML, unproven singularity theories, proven data science success stories, lurking cybersecurity dangers, and reimagining the digital experience. You can reach him at [email protected].
Image credit: iStockphoto/liveostockimages