Data steward role needs some shepherding itself

Investments in data governance are largely motivated by a desire to improve business operations and performance through better oversight and management of corporate information. But while a data governance program institutes policies and processes designed to produce more accurate and consistent data throughout an organization, it primarily becomes the job of the data steward to put those policies and processes into practice by ensuring compliance with them. To a large degree, the success of a data governance strategy depends on the success of its associated data stewardship efforts.

It's important to set up a data governance management structure and operating model in a way that is aligned with an organization's existing structure. That includes the data stewardship element. But doing so isn't as easy as it sounds, nor is managing the work of data stewards.

One of the issues that most frequently stymie data governance initiatives occurs when the urgency to launch a program leads to decisions that may be counterproductive. In one example, a company selected a pool of data stewards before defining what they were supposed to do, resulting in significant confusion. In another, the need to demonstrate that progress was being made quickly led an organization's data stewards to embark on an undirected metadata survey involving hours of pointless tasks.

As a counterpoint to those unsuccessful approaches, here are seven suggestions on how to effectively structure and manage a data stewardship team that can help keep a governance initiative in line:

Formalize the position. Prior to asking individuals to take on the data steward role, ensure there is a formal enumeration of its responsibilities; the skills needed to fulfill it; metrics for measuring performance; and details about how the stewardship duties would mesh with existing job responsibilities, assuming that it wouldn't be a full-time position.

Get granular on stewardship roles. There really are a number of different roles that could warrant the title "data steward" -- a metadata steward and an operational data steward, for example. It's best to clearly delineate what distinguishes these roles from one another, as well as how the employees filling them will work together to support the data stewardship process.

Establish business ownership of data. Data stewards may be responsible for aspects of compliance with a data governance policy, but that doesn't mean they're accountable for the data itself. Ownership and accountability must remain with the appropriate business unit or department.

Get aligned with the business. As part of a data governance program, data usability expectations are framed within the context of expected business improvements, such as increased revenue, decreased costs, reduced risks and enhanced productivity. But most IT and data management practitioners are more familiar with the mechanics of managing data than they are with the corresponding business processes in which it's used. If your data stewards don't come from the business side themselves, align them with key subject-matter experts in business units to help identify data issues and prioritize remediation tasks.

Create incentives. As opposed to a typical project in which there are well-defined milestones and deliverables, the nature of data stewardship is to ensure against the occurrence of data incidents. Develop an incentive program for your data stewards that recognizes and rewards them for their ability to meet performance objectives.

Find people with the right skills. Because the data steward role is still evolving, advertising for individuals with years of experience might not yield a bumper crop of good candidates. And in many companies, data stewardship isn't a full-time occupation. As a result, you might have to cast a net inside for people with stewardship potential. Consider which data management skills are a necessity -- but more important, look for workers with valuable, transferable skills, such as good communication capabilities, assertiveness in promoting best practices and comfort with the idea of acting as agents of change.

Give data stewards appropriate tools. Although data stewardship is fundamentally a procedural matter, there are some tools that can support stewardship initiatives, including data quality assessment, data validation, and data incident reporting and management software -- perhaps even a data quality and data stewardship scorecard application.

All these steps share a common theme: They're predicated on investing a reasonable level of effort up front in designing the structure of a data governance and stewardship program, as well as the operational processes for making it work. Once that's done, hiring the right people, placing them in roles that are clearly defined, getting them in sync with business units and motivating them with performance incentives will help enable a practical and sustainable data stewardship process.

About the author:
David Loshin is president of Knowledge Integrity Inc., a consulting, training and development services company that works with clients on big data, data quality, data governance, master data management and business intelligence projects. He also is the author of four books, including The Practitioner's Guide to Data Quality Improvement and Master Data Management. Email him at [email protected].

Email us at [email protected] and follow us on Twitter: @sDataManagement.