While it’s risky to generalize, I can say with near certainty that you do need such a position — and not just one, but probably several or even many. The actual number you need is correlated to the number of business data elements (BDEs) in your master BDE inventory and the total number of physical instances associated with those BDEs.
Before we discuss the value of data stewards, however, a quick definition is in order. A data steward is an individual employee selected by an organization to manage the way a specific BDE (for example, “customer name”) is captured and used through the enterprise’s various spreadsheets and databases. Since an organization can have hundreds of BDEs, the number of data stewards needed can be large or small, depending on the span of control and authority a data steward is allowed to have.
The easiest answer for why you need data stewards is because when managed effectively, their efforts are crucial to increasing revenues, proactively managing business risk and/or reducing costs. Data stewards can help increase revenue by enhancing your organization’s ability to:
Analyze product/service quality and time-to-market
Improve decision making
Enhance customer/market retention
Looking at it from another direction, data stewards can help reduce costs in such areas as:
Core business processes and technology
Management and/or elimination of redundant data
Perhaps the broadest value of the data steward role is reducing risk. Some of this risk may come in the form of compliance with data reporting requirements in such laws as Sarbanes-Oxley and Dodd-Frank. Other risks may come in the form of hits to public opinion or market perception, should you accidentally release inaccurate data. In either case, a failure can have catastrophic consequences. Fortunately, effective data stewardship can help you avoid doing so — or at least, greatly minimize the likelihood.
Finding the right data stewards is worth the time
To appreciate the data steward’s value, consider how the role is managed in different organizations. In general, the duties of the role are fairly specific: to manage and curate one or more data elements for the organization. But beyond that, trying to describe generic job requirements is close to impossible.
In practice, organizations wrestle with how to fill this role, and how to manage the people in it. Companies generally take one of three of approaches to appointing data stewards — only one of which is recommended.
Some organizations assign data stewards based on the subject area of the data involved, which sounds reasonable enough on first blush. In an insurance company, for example, many BDEs have to do with policies — policy number, effective date, issue date, etc. — so responsibility for them might be assigned to employees in the billing function. But this approach tends to fall short, because these subject areas often overlap with others, such as name/customer name, and so on. In our experience, this approach creates a lot of needless confusion; what’s more, one can never make the boundaries distinct enough for the individual data steward to be successful.
Another widespread way to assign data stewards is by functional area within the organization. For example, accounting might be asked to designate one of its staff members as a data steward for a specific financial BDE; meanwhile, customer service is directed to name one of its team as data steward for a BDE that tracks customer inquiry status, and so on throughout the enterprise. Again, it makes a certain kind of sense to spread the responsibility around the organization. The problem is that doing so can create some awkward overlaps, redundancies and gaps. For example, the BDE for “customer name” probably appears in dozens of columns and databases throughout the enterprise, so choosing the best department or individual to “own” it becomes its own challenge. But the most important failing of this approach is that it probably won’t result in finding the individuals who actually know the most about a given BDE, or who have the most “skin in the game.”
In our experience, the best and only way to assign data stewards is to start by building an inventory of data elements — a record of all the hundreds of distinct data elements as they occur throughout the organization’s information systems. The next steps are to implement a process for determining how the data is created, used, and transformed as it moves through the organization, and then enriching the data elements and finding anomalies. When this process is managed well, the individuals who know the most about each data element emerge naturally, making it a straightforward decision as to whom to assign as data stewards for specific elements.
The challenge of this approach is that it requires a fairly mature organizational machine. The organization must be able to comb through its data landscape, seek out and sort through all the believable experts for a given area of data and engage in aggressive, rigorous collaboration processes to extract the knowledge from employees’ heads that contributes to understanding the organization’s data. Typically, organizations want to take a shortcut — assigning the work described above to a group of “named” data stewards and “holding them accountable” to deliver. This approach is ineffective, as data stewards on day one are usually unaware of everything they need to know in order to succeed.
The big lesson: stop “assigning” data stewards — start thinking about cultivating them instead.
It’s important to note that even before you start on the process of identifying and cultivating data stewards, your organization should strongly consider establishing an Office of Data (led by a Chief Data Officer, or CDO). This office represents a new type of function within your organization.
In addition to overseeing the process of building a data element inventory and determining how data is created, used, and transformed in the organization, the Office should be tasked with data governance, metadata management, data quality and data architecture —responsibilities that today might reside across multiple functions within your organization.
Data stewards help to leverage a strategic insight that should already be central to your organization: that data is valuable, ubiquitous, and vulnerable. Managed effectively by the Office of Data, the role is performed by the people who are closest to and most knowledgeable about the data — and therefore in the best position to ensure its integrity.