As a data-driven organization in the modern, hyper-competetive business landscape, it’s imperative that employees, business leaders and decision makers can understand your data.
In a previous article, I argued that business process management without data governance is a perilous experiment. The same can be said for enterprise architecture initiatives that traditionally stop at the process level with disregard for the data element.
Therefore, an organization that ignores the data layer of both its process and enterprise architectures isn’t tapping into their full potential. You run the risk of being siloed, confined to traditional and qualitative structures that will make improvement and automation more difficult, time-consuming and ultimately ineffective. However, it does not have to be this way.
I’m going to make a bold statement, and then we can explore it together. Ready? Without data governance, a process management or enterprise architecture initiative will result in a limited enterprise architecture and any efforts that may stem from it (process improvement, consolidation, automation, etc.) also will be limited.
So how exactly does data governance fit within the larger world of enterprise architecture, and why is it so critical?
A constant source of unpleasant surprise for most medium and large-scale enterprise architecture and process management initiatives is discovering just how tricky it is to establish interconnectivity between low-level processes and procedures in a way that is easy sustainable and above all, objective.
Traditionally, most projects focus on some type of top-down classification, using either organizational or capability-based groupings. These structures are usually qualitative or derived from process owners, subject matter experts (SMEs) or even department heads and business analysts. While usually accurate, they are primarily isolated, top-down views contained within organizational silos.
But more and more enterprise architecture initiatives are attempting to move beyond these types of groupings to create horizontal, interconnected processes. To do so, process owners must rely on handover points – inputs and outputs, documents and, you guessed it, data. The issue is that these handover points are still qualitative and unsustainable in the long term, which is where data and data governance comes in.
By providing an automated, fully integrated view of data ownership, lineage and interconnectivity, data governance serves as the missing link between disparate and disconnected processes in a way that has always proved elusive and time consuming. It is an objective link, driven by factual relationships that brings everything together.
Data governance also provides an unparalleled level of process connectivity, so organizations can see how processes truly interact with each other, across any type of organizational silo, enabling realistic and objective impact analysis. How is this possible? By conducting data governance in conjunction with any enterprise architecture initiative, using both a clear methodology and specialized system.
Everyone knows that processes generate, use and own data. The problem is understanding what “process data” is and how to incorporate that information into standard business process management.
Most process owners, SMEs and business analysts view data as a collection of information, usually in terms of documents and data groups such as “customer information” or “request details,” that is generated and completed through a series of processes or process steps. Links between documents/data groups and processes are assumed to be communicated to other processes that use them and so on. But this picture is not accurate.
Most of the time, a document or data group is not processed in its entirety by any subsequent/recipient processes; some information is processed by one process while the remainder is reserved for another or is entirely useless. This means that only single, one-dimensional connections are made, ignoring derived or more complex connections.
Therefore, any attempted impact analysis would ignore additional dimensions, which account for most of the process improvement and re-engineering projects that either fail or present significant overruns in terms of both time and budget.
With data governance, data is identified and tracked with ownership, lineage and use established and associated with the appropriate process elements, showing the connections between processes at the most practical informational level.
In addition and possibly most important, process ownership/responsibility becomes more objective and clear because it can be determined according to who owns/is responsible for the information the process generates and uses – as opposed to the more arbitrary/qualitative assignment that tends to be the norm. RACI and CRUD matrix analyses become faster, more efficient and infinitely more effective, and for the first time, they encompass elements of derived ownership (through data lineage).
Process automation projects also become more efficient, effective and shorter in duration because the right data is known from the beginning, process interconnectivity is mapped objectively, and responsibilities are clearly visible from the initial design phase.
With requirements accompanied by realistic process mapping information, development of workflows is easier, faster and less prone to errors, making process automation more attractive and feasible, even to smaller organizations for which even the scoping and requirements-gathering exercise has traditionally proved too complicated.
Data governance enables an organization to manage and mitigate data risks, protecting itself from legal and reputational challenges to ensure continued operation. And once data is connected with business processes through effective, proactive data governance, additional benefits are realized:
While the above benefits may appear vague and far-removed from either a pure enterprise architecture or data governance initiative, they are more tangible and achievable than ever before as organizations begin to rely more on object-oriented management systems.
Combining the strategic, informational-level detail and management provided by data governance with the more functional, organizational view of enterprise architecture proves that one plus one really does equal more than two.
To learn more about how data governance underpins organization’s data strategies click here.