Data stewardship (DS) is the practice of overseeing an organization’s data assets to ensure they are accessible, reliable, and secure throughout their lifecycles. It is a framework of roles, responsibilities, and processes designed to support the organizational strategy through a data governance (DG) program.
At its core, data stewardship comprises data stewards – formalized roles that take responsibility for their data. These people:
Manage data so that it meets the needs of the organization, by supporting the data strategy.
Facilitate collaboration between IT and business units.
Work with DataOps to improve information flow and do technical maintenance.
Support a data quality program to meet compliance and accessibility needs.
Implement operational policies, rules, and procedures issued by data governance.
Promote a data-driven culture through data literacy.
These activities necessitate stewardship as a continuous presence within the organization, capable of evolving with the business. Consequently, data stewardship may prioritize certain components over others at various times.
Data stewardship encompasses many different aspects of data management. It can be described through these perspectives:
Process Oriented: This approach characterizes data stewardship from end to end.
University of Michigan: Data stewardship is the means to “encourage desirable behavior in the valuation, creation, use, storage, archiving, and deletion of information.”
Airbyte: Data stewardship focuses on data management throughout the data “lifecycle – from creation to storage and deletion.”
Strategic: These definitions focus on stewardship benefits for companies.
DATAVERSITY® article: Data stewardship consists of stewards who care for data to get better business insights and make more informed decisions.
The Data Management Body of Knowledge: Data stewardship helps “organizations get value from its data.”
Data Quality: Some sources focus on data stewardship’s promise to keep data usable and fit through a high data quality.
TechRepublicarticle: DS ensures that an organization’s data assets are accurate, consistent, and compliant with all relevant regulations.
Data stewardship covers several aspects of data, and organizations need different types of data stewards.
Types of Data Stewards
Different data stewards perform a wide range of tasks. Organizations assign various roles to care for data to get the work done efficiently.
Ideally, businesses recognize that everyone working and servicing data is a steward. Also, companies need to have more than one data steward for each data type. For example, a customer service account manager takes the critical business pieces, and IT ensures the person’s login works properly.
Companies determine stewardship needs based on the culture and mission objectives. Here are four possible ways to tailor stewardship roles:
Expertise: Organizationsassign people as data stewards according to their strengths – e.g., knowledge keepers are subject matter experts (SME) with insider “tribal” knowledge. These data stewards represent their teams, collaborate across departments, and often train others on their specialized knowledge.
Level of Responsibility: Across the organization, different roles handle decision-making according to their level of responsibility. For example, data definers, producers, and users do data operations. These members would take on data profiling and monitoring, updating data documentation, and resolving problems.
Activity Types: Businesses can divide data stewards with activity types. A technical data steward focuses on coding operations, like system maintenance, while a business data steward defines the data in a glossary or catalog.
Technology: Companies may tie data stewardship roles to a particular system. For instance, a company with a master data management (MDM) data system may assign different roles based on its data schema. This could include customer, product, location, or organizational hierarchical data.
Thus, various people divide stewardship tasks according to business needs. How they do so will depend on the organizational strategy.
Why Is Data Stewardship Important?
Many organizational strategies engage stewardship to break down data silos, a major obstacle to leveraging data assets effectively and efficiently. According to ITPro, 81% of IT decision-makers saw them as the biggest barrier.
Data stewardship bridges data across the organization through better communication between technology and business units. Consequently, it better achieves organizational goals through data usage and increases data literacy. Organizations are able to better do the following:
Remediate data and data-related issues and problems
Operationalize data governance
Achieving these stewardship benefits depends on how organizations formalize and structure them through their data governance programs.
Is Data Stewardship the Same as Data Governance?
Data stewardship plays a key part in data governance. Stewards implement DG policies, rules, and procedures, in alignment with the enterprise’s strategy.
However, data stewardship responsibilities also span beyond data governance. They involve these data management activities:
Monitoring, identifying, and rectifying data quality issues
Improving organizational data literacy so all the departments can understand and use data well
Facilitating communication and collaboration with other teams and departments
Fostering a data-driven culture through a curious mindset about data
Likewise, data governance responsibilities also go beyond data stewardship. For example, a DG body may decide on a comprehensive technical and procedural fix to a data quality issue.
Consequently, data stewardship needs to be clearly defined and communicated across the organization. Then its overlap with data governance becomes decipherable.
Challenges to Data Stewardship
Data stewardship can be challenging without clear direction and agreement across the organization. For example, Shaw Industries had a healthy staff of data stewards, but no coherent data governance framework. Furthermore, its data managers were resistant to changing from old pre-electronic processes.
Consequently, although the company had data stewards, they continued to have siloed data sets and segmented attention to them. Shaw Industries needed to change the mindset so everyone was on the same page on what to do.
This kind of mandate to change to a data-driven culture makes stewardship one of the most challenging initiatives to implement. According to a 2024 Gartner Chief Data and Analytics Officer Agenda Survey, only 43% of respondents had successful stewardship efforts.
Companies find additional difficulties in adequate stewardship, including:
Role Assignment: Hand-picking the stewards leaves gaps because some people who use that data will not consider themselves accountable. Moreover, workers resent more work, especially without additional pay.
Too Much Stewardship: Too many stewards will lead to confusion. Different people will have varying interpretations of the organization’s data governance policy and disagree. Otherwise, people will spend time reinventing the wheel, when a good stewardship process is already in place.
Misunderstandings: IT thinks data is a business problem, while business thinks IT is managing data adequately. However, both groups misunderstand who is taking care of what, and how.
Coverage: Organizations need to prioritize whether the data should be stewarded, because there is just not enough money or time to ensure every piece of existing data has a caretaker. Furthermore, not every data set may require coverage, for example the travel plans for a company retreat.
Despite these challenges, a good stewardship program can work well and is essential. The future requires that organizations work through their stewardship challenges to leverage new technologies.
The Future of Data Stewardship and Its Role in Emerging Technologies
Data stewardship will become increasingly important and more complex and evolved in the future. This need is underlined by emerging technologies and increasing regulations.
Artificial Intelligence (AI) and Data Stewardship
As companies increase their adoption of AI, data stewardship will benefit by:
Fostering Awareness of Data Challenges: AI will make it easier to sort, assess, categorize, and triage different data issues. Using generative AI, which generates content, companies will see connections among various pieces of data. They will get new insights on solving problems and better present these findings to stakeholders.
Classifying Data with Sensitivity Labels: AIalgorithms can automatically identify and classify data according to business metadata standards. This advantage improves steward assessment, by categorizing data more efficiently and effectively for compliance and quality remediation.
Cultivating Regulatory Transparency: AI solutions can digest federal and regulatory details and flag issues. Additionally, they can assist in locating the questionable information suggesting a resolution.
Showcasing Program Value: AI connects relationships among people, processes, data, and technology. Data stewards can target this information, providing evidence for the value of stewardship.
Although advancing AI will streamline and benefit stewardship activities, it will also require more data stewardship attention. In addition to humans, AI generates, transforms, and consumes data. So, data stewardship will need to handle all these situations, in addition to what human-touch data has and will continue to do.
Real-Time Decision-Making
Real-time data stewardship will become increasingly important, especially in ensuring compliance and adequate data quality.
Many systems stream data. As companies ingest and process this data, analysts face the challenge of instantaneous analysis and timely response.
For example, as the Internet of Things (IoT) continues evolving, stewards will be required to ensure good data quality, compliance, and immediate resolution of any issues. That way, these devices can better function, identifying anomalies and predicting maintenance.
Ethics and Laws
The number of data and AI legal requirements will continue to grow. As of this writing, there are more than 120 AI bills in the U.S. Congress.
Organizations will need to keep up with this legislation by adding and reviewing data stewardship protocols and activities. These upcoming bills will also encourage corporations to pay more attention to security and privacy.
Cross-Corporate Collaboration
In addition to cross-functional collaboration across an organization, data stewardship will need to consider cross-corporate collaboration among many organizations. Businesses will be encouraged to work together or join data trusts to share some data assets due to demands for data to fuel AI and a desire to maximize limited resources.
This shift will require trusted third parties (an intermediary organization or agreed upon representatives) to take responsibility for stewarding and governing common data sets that all these companies can access.
These intermediaries already exist. For example, The Health Care Cost Institute (HCCI) cares for data from insurance companies in the United States (e.g. Aetna, Humana, Kaiser Permanente, and United Healthcare). It shares health care and cost information with researchers, but removes identifiers about which company has provided the data before sharing them.
In the next couple of years, these cross-corporate alliances will continue to grow across industries. As they do, data stewardship will continue to expand and provide additional services to cover these needs.
