Implementing Data Governance at Transportation Agencies: Volume 1: Implementation Guide (2024)
Chapter: 4 Roles and Responsibilities
4. Roles and Responsibilities
Overview
Establishing roles and responsibilities is an essential part of data governance. As discussed in Chapter 3, there can be different levels of formality and each agency should tailor the approach to their situation and needs. This chapter provides guidance on common DOT models for staffing a data governance function, setting up decision-making and advisory bodies and defining data stewardship roles.
Data Governance Leadership and Staffing
It is difficult to make meaningful changes to how data are managed in an agency without a data governance lead and at least one full-time staff person (or equivalent across multiple individuals and consultants) to support the lead. Adding staff and consulting resources to this core team will accelerate progress.
DOT Data Governance Leads Have These Titles:
- Chief Data Officer
- Data Governance Officer
- Enterprise Data Manager
- Geospatial Data Officer
- Information Governance Coordinator
The Data Governance Lead. Designating a lead individual is the first step in the data governance implementation covered in Chapter 3. This initial lead can be an interim responsibility while the agency is determining its approach to data governance. However, any agency hoping to realize the benefits of data governance discussed in Chapter 2 will need a permanent data governance lead, responsible for strategy, management, and oversight of data governance activities. This should be a full-time data-focused position, though their portfolio can extend beyond data governance into data program management, reporting and analytics.
DOT Data Governance Leads Report to:
- Chief Administrative Officer
- Chief Information Officer/IT Director
- Planning Director
- Operations Director
- Chief Engineer
It generally works best if the lead reports to someone on the agency’s leadership team (e.g., a deputy, chief or division director) because this maximizes the chances that data governance will have the level of sponsorship and support needed to succeed. DOTs have used a mix of strategies for establishing data governance leads, including designating an existing employee, recruiting outside talent with data governance experience, and hiring a full-time consultant to play the role and mentor internal staff to take this on after 1-2 years.
See Chapter 3 for the desirable qualities of a data governance lead
See Chapter 7 for sample competencies for an agency data lead
Location of the Data Governance Lead. Data governance leads at DOTs are located within a variety of different functional areas (see sidebar). Many DOTs have embedded data governance within their planning, engineering, operations, performance management, or research areas and tied them to major data and/or analytics programs.
Some DOTs have located data governance within their IT unit. This can be advantageous for establishing a strong linkage between data and IT governance. While a close relationship between IT and data governance is essential, it is important to recognize that the scope of data governance intersects with but goes far beyond information systems development and acquisition processes. IT should be viewed as an enabler of data governance rather than a driver of data governance. If the data governance lead is housed within the IT unit, they should avoid getting overly focused on technology and system development and have a clear data-centric (rather than application-centric) perspective.
Another option that some agencies have used is to place data governance parallel to the IT unit – so that they both report to the same manager (e.g., the Chief Administrative Officer).
Data Governance Support Roles. The data governance lead will require assistance from agency staff and/or a consultant team to do more than coordinate data governance meetings. The level of required staffing will depend on the target goals and accomplishments. However, a minimum of one full-time equivalent (FTE) to support the lead is recommended. One common model is to create a Chief Data Steward role to serve as the lead’s deputy or assistant and take on responsibility for working with business units to support data governance implementation activities. This takes an operational burden off of the data governance lead, enabling them to focus on strategic issues, resourcing, communication and liaison with the management team and external parties (where applicable).
Engage staff who might serve in support roles early on to make sure their perspectives are heard - see the discussion of stakeholder engagement as part of goal-setting in Chapter 3.
Support roles should report to the data governance lead or assistant/deputy. However, it may also be possible to tap into existing staff within other parts of the agency – for example, the communications office; IT group; or units responsible for research, library services, organizational improvement and performance management. They may be willing to provide support on a part time basis or play a role in limited term initiatives.
Table 2 presents a set of support roles and provides ideas for existing agency staff that might be tapped to perform these roles. Agencies can use this table as a template for identifying resources to leverage.
Table 2. Data Governance Support Roles
| Role | Existing Staff to Leverage |
|---|---|
| Data Librarian or Metadata Specialist – build and manage data inventories, catalogs, and glossaries Desired Qualities: detail orientation, organized, library or information science background preferred. |
Librarian (DOT or partner University) Research Office Staff Data Analyst Data Architect GIS Staff Data Warehouse Staff |
| Data Governance Generalist – provide a variety of support to the data governance lead and assistant/deputy; work directly with business units to conduct training and help them to follow the guidance and standards established by governance (for example, providing standard metadata). Desired Qualities: people skills, facilitation skills, good communicator of technical information, understanding of DOT business and culture, teaching experience, experience producing and/or using data |
Data Analyst Data Architect GIS Staff Data Program Manager (e.g., traffic, road inventory, survey, right of way) Data Warehouse Staff Planner/Analyst |
| Data Quality Specialist – develop and support data quality management processes and tools. Desired Qualities: people skills, good communicator of technical information, understanding of DOT business and culture, teaching experience, experience producing and/or using data |
Data Analyst Performance Management Staff GIS Staff Data Warehouse Staff |
| Communication Specialist – develop communication materials (website, fact sheets, presentations) and coordinate stakeholder engagement activities. Desired Qualities: communications, graphic design, facilitation |
Public Affairs Unit Staff Graphic Artist Planner/Analyst External communications/public relations consultants |
| Change Management Specialist – create and carry out a change management plan (can also cover the communications responsibilities above if the right resource is identified with both skills). Desired Qualities: organizational change management training, strategic thinking, understanding of organizational dynamics and culture, interpersonal skills, project management experience |
Process Improvement or Organizational Change Management function staff External change management consultants |
| Data Architect – map existing and target data architectures; identify and pursue opportunities to move towards target architectures and increase data standardization and interoperability. Desired Qualities: computer science or information science degree, data modelling expertise, analytical thinker, communication skills, familiarity with agency data and business processes |
Information Technology staff with background/training in data architecture. Business Analysts External data architecture consultants |
| Role | Existing Staff to Leverage |
|---|---|
| Data Security and Privacy Specialist – develop and oversee application of procedures for classifying private and sensitive data. Work with technology partners to ensure data protection. Desired Qualities: understanding of data security and privacy risks and mitigation strategies, communications skills, familiarity with agency data and business processes |
Information Technology staff currently charged with data security responsibilities Data Privacy Office staff Chief Counsel’s Office staff |
Data Governance Bodies
If your agency has chosen to initially pursue a decentralized operating model for data governance (see Chapter 3, Step 6, option 3), you can skip this section, since you will be coordinating data governance out of the operating unit that houses the data governance team and relying on existing management groups to provide additional coordination and decision-making as needed.
If you have chosen a centralized operating model for data governance (see Chapter 3, Step 6, options 1 or 2), you will need to set up one or more data governance bodies. This is an important step, which if not approached correctly can result in a stalled data governance effort. Keep in mind that data governance bodies serve several important purposes:
- Alignment. Data governance bodies keep efforts focused on adding value and avoiding unnecessary new rules or processes. They work to align data governance with agency business needs, priorities and other initiatives. They provide the perspective needed to design governance processes that fit with the agency’s culture and resource constraints.
- Legitimacy. Data governance bodies are empowered by the agency’s leadership to make decisions that “stick”. They can adopt new data policies and standards and have the authority to follow through and support their adoption.
- Engagement. Data governance bodies enable connections and communication channels across the agency that are essential for making change. They involve key stakeholders in shaping and prioritizing data governance initiatives and establish the scaffolding on which to build broad engagement from data producers, managers, and consumers throughout the agency.
Composing an Effective Data Governance Council
- Breadth – include representatives of key data programs and organizational areas that will be impacted by data governance.
- Motivation – include people who want to see data improvements and are willing to participate and serve as change agents.
- Expertise – include people who understand the agency’s major data programs and information systems.
- Authority – include people who have the positional or relational authority to enforce adherence to that new policies or standards.
Primary Data Governance Body. Most organizations set up a primary data governance body, which may be called a Data Governance Committee, Council or Board. This group’s purpose is to guide development and adoption of new policies, standards, guidance, and/or procedures. Its membership will be similar to the stakeholder group discussed in Chapter 3, Step 2 – but it may be
necessary to reconstitute this group to provide greater organizational reach, authority, and engagement.
Executive Oversight Group. To provide continuing top management support and engagement, consider establishing an executive-level group that provides strategic direction and serves as the final authority for resolving particularly difficult issues that cannot be handled by lower-level groups. If there is an existing executive team that meets regularly, there may not be a need to create a separate group to serve this purpose.
Data Stewards Committee. To provide a bridge between the primary data governance group and the people responsible for managing and stewarding data, consider creating an operational-level group. This group can be used to provide “on the ground” feedback on proposed standards and processes and can facilitate their implementation. An alternative approach is to establish working groups on an as-needed basis to support specific data governance initiatives.
Formalizing Data Governance
Some agencies have found it difficult to sustain data governance and achieve meaninful change without documenting decision authorities, roles and responsibilities in a formal manner.
This means adopting an agency data governance policy, establishing charters for each data governance body, and creating a RACI chart (or similar) that spells out roles that are responsible, accountable, consulted and informed for each major data-related decision.
Table 3 outlines typical members and responsibilities for these three levels.
Table 3. Data Governance Bodies: Members and Responsibilities
| Data Governance Group | Members | Example Responsibilities |
|---|---|---|
| Executive Oversight Group | Executives – direct reports to the agency CEO or Deputy Director | Provide strategic oversight for data governance to align with agency priorities – approve mission, vision, and goals statements. Approve changes to agency policy. Resolve issues escalated by the Data Governance Committee/Council/Board. |
| Data Governance Committee/Council/Board | Senior Managers and technical leads from data programs, data-intensive business units, and IT | Define and periodically review mission, vision, and goals for data governance. Review and prioritize data governance actions/initiatives. Formally adopt new policies, standards, guidance, and procedures. Advise on data governance initiative design (typically through work groups). Identify opportunities for coordination and collaboration. Flag issues and concerns that may impact success of data governance initiatives. Serve as liaisons to external groups (e.g., state-level geospatial data committee or Traffic Records Coordinating Committee). Resolve issues brought to them by stewards and others. |
| Data Governance Group | Members | Example Responsibilities |
|---|---|---|
| Data Stewards Committee | Business line or subject area data representatives; can also include district or regional office liaisons | Review and comment on proposed guidance, standards, procedures, and policies. Communicate with data stewards and others to disseminate information about data governance initiatives and provide feedback. Identify issues, challenges and opportunities for coordination related to data governance implementation. Actively participate in data governance implementation activities through pilots and training activities. |
Tips for Maximizing Effectiveness of Governance Bodies
- Engage members in creating a charter to build agreement on purpose, role and ground rules.
- Create an annual schedule of regular meetings.
- Plan agendas to include a mix of informational items and items requiring decisions or input.
- Build ownership by designating chairs or co-chairs and assigning responsibility for agenda items.
- Use a skilled facilitator to keep meetings on time and on task and to maximize participation.
- Follow up with members who don’t attend meetings and encourage them to attend or send an alternate.
- Document decisions and action items.
Maximizing the Effectiveness of Data Governance Bodies
Once data governance bodies are established, they need to be skillfully managed to maximize their value. Some agencies have struggled with attendance and productive engagement on the part of data governance body members. The data governance team plays an important role in making these data governance bodies successful by conducting background work, framing up issues and decisions, making recommendations and helping the governance body members understand what they are being asked to do – and the pros and cons of different courses of action. Specific practices that can be used to make sure that members of governing bodies stay engaged and accomplish their intended purposes are summarized below and highlighted in the sidebar.
Resourcing. It is important to allocate sufficient resources to support whatever governance bodies are established. Support is needed for member recruitment, onboarding and replacement, agenda planning, meeting scheduling and facilitation, materials preparation, and notes.
Regular Meetings. With busy practitioners, it is often preferable to have a regularly scheduled meeting at a predictable frequency than to try to schedule meetings “as needed.” The added logistical work of scheduling each time one is needed adds cost to the data governance effort. The frequency of meetings can be adjusted if it turns out that the originally established schedule is insufficient or too frequent. Canceling an already-scheduled meeting due to a lack of content is preferable to try to get on peoples’ calendars without a long lead time.
Work Groups and Subcommittees. Data governance bodies can form ad-hoc work groups and standing subcommittees to deal with specific issues that can best be addressed by a sub-group with domain expertise. These sub-groups can report back to the parent group with their findings and recommendations. This
See Chapter 7 for a sample Data Governance Body Charter outline.
can work well for topics like how to measure data quality performance, metadata requirements, spatial data standards, and domain-specific data standards (e.g., traffic monitoring for two-lane rural roads).
Advisors. There may be experts on specific topic areas that do not need to be official members of the Data Governance group. They can be invited on an as-needed basis based on the agenda or receive a standing invitation to participate in an advisory role as they see fit.
Data Stewardship Roles
Data stewardship means assigning business responsibilities for classification, protection, quality, and effective use of an organization’s data assets. Strong data stewardship is integral to successful data governance because stewards are the people that make sure the standards and guidelines established by governance are actually followed.
It is a relatively straightforward task to define data stewardship roles and responsibilities – this has been accomplished by several DOTs. It is more challenging to identify people to play these roles and keep up with the inevitable employee turnover. Perhaps the biggest challenge of all is to make sure that the designated stewards are actually carrying out the defined responsibilities and hold them accountable.
“The entire data governance framework is essentially the vessel for data stewardship.”
- John Ladley, data governance consultant and author [4]
Introducing New Responsibilities for Data Stewardship
- Low Change Approach: Identify individuals already playing stewardship roles and offer tools and support to help them improve.
- Medium Change Approach: develop a standard description of data steward roles and responsibilities and implement a process for designating data stewards.
- High Change Approach: Add specific language to job descriptions with responsibilities and expected time commitments for individuals who serve as Data Stewards.
Defining data stewardship roles
Table 4 shows different types of data stewardship roles that have been defined by transportation agencies. These can be used as a starting point to consider what roles to establish in your agency.
Table 4. Data Stewardship Roles
| Role | Description | Example Functions |
|---|---|---|
| Data Owner or Trustee | Individual with decision-making authority and responsibility for data related to a particular subject or program | Identify and advance initiatives to improve the value of data within their scope Work with IT staff to protect and control access to data within their scope (e.g., approve access requests, identify essential data to prioritize for disaster recovery). |
| Business Data Steward (other names include Operational Data Steward, Subject Area Steward) | A subject matter expert (SME) with designated responsibilities for classification, protection, quality, and effective use of one or more data sets | Define data needs within their functional areas. Understand and answer questions about the meaning, derivation, quality requirements, and uses of data within their scope. Define business rules and data quality objectives. Work to enhance the quality and value of data to the organization. Document data within their scope. |
| Role | Description | Example Functions |
|---|---|---|
| Data Custodian (sometimes called Technical Data Steward) | An individual with physical custody of the data, typically an information technology professional with skills needed to perform data processing tasks. | Assist with technical data documentation. Perform data manipulation including loading, transfer, transformation, and integration. Implement data confidentiality and security requirements including access controls. Perform or monitor data backup and recovery processes. Create and run data profiling and validation scripts; assist with data quality defect resolution. |
| Coordinating Data Steward (other names include Data Domain Steward and Enterprise Data Steward) | Designated point-person for data within a particular organizational unit, function, data domain or subject area. Responsible for facilitating data improvements and issue resolution across business units. Often serves as a liaison to the data governance body. | Identify and communicate data needs, opportunities, and issues for their domain to be addressed through data governance. Coordinate implementation of data standardization and improvement efforts within a major organizational unit (e.g., division or branch). Identify and support business data or subject area stewards within their designated area. Facilitate cross-divisional collaboration on data improvements. Draft and review proposed new policies, standards, processes, and guidelines. Resolve cross-cutting data issues that go beyond an individual business unit or system. Support data governance compliance. Advocate for data quality and documentation initiatives. |
Owner or Trustee?
Some agencies choose not to use the term “data owner” because it reinforces the notion that individual employees own data (as opposed to the agency or the state or the citizens). Trustee is an alternative that connotes someone entrusted with managing an asset for the benefit of others.
Existing Roles. The first three roles: Data Owner/Trustee, Business Data or Subject Area Steward and Data Custodian are roles that already exist in your agency – i.e., there are already people who oversee and manage data. They just need to be identified and formally recognized. This will allow you to:
- Clearly establish accountability and responsibility for data management and decision-making;
- Target data management guidance and training to the right people;
- Identify who needs to be engaged in efforts to standardize particular data elements across data sources; and
- Include contact information for particular data assets in your data catalog or inventory.
Formalizing data stewardship responsibilities
Florida DOT has updated duty statements to reinforce data stewardship roles. Caltrans provides standard language for different roles that hiring managers can use, including a suggested range for percentage of time on data stewardship activities. be used and includes a suggested range in percentage of time.
Some individuals who are designated as data stewards may feel that they do not have sufficient time to take on any new responsibilities that the data governance body determines to be an agency priority (e.g., for preparing metadata). In the short term, this issue can be addressed through providing
support to the data stewards to help them fulfill any new or modified responsibilities. In the longer term, data stewardship responsibilities can be more explicitly spelled out in position descriptions and recognized as integral to the job. There may also be opportunities to combine related stewardship or liaison type roles with responsibility for records management, document management, and web content management.
Coordinating Data Steward Role. The last role of Coordinating Data Steward is new. It establishes a middle-level steward that sits in between the data governance body and individuals serving in other stewardship roles. This middle level role is helpful, particularly in large DOTs where there may be one hundred or more business data or subject area stewards and data owners/trustees. The Coordinating Data Stewards can meet as a group and provide a conduit for guidance, information and feedback between the data governance body and the larger group of stewards. While establishing this role should not require a new hire, it will involve a time commitment on the order of 1-3 days per month (or more) for an existing employee.
Relationship between data and IT Roles. Transportation agency IT departments typically designate “system owners” to be the primary business points of contact for application development and enhancement decisions. In many cases, business system owners will be the same people as data owners/trustees or stewards. However, this Is not always the case. For example, some systems may have multiple modules with different types of data (e.g. an enterprise asset management system). There may also be a single data owner or steward for a data domain that appears across multiple systems (e.g. road inventory data), in which case there would be multiple systems owners but a single data owner.
Coordinating Steward Roles at DOTs
Minnesota DOT has nine data Domain Stewards who are accountable for managing shared data for the common benefit of the agency.
Florida DOT has designated Enterprise Data Stewards representing 20+ functional areas (e.g. planning, design, safety). They provide leadership and guidance for data stewards and custodians, and elevate needs and suggested data improvement initiatives to the executive level data governance group.
Virginia DOT initially hired a single, full-time Enterprise Data Steward to coordinate ~80 Data Stewards across all agency data areas. However, they ultimately anticipate designating eight full-time Enterprise Data Stewards, one assigned to each of their highest-level data domains.
Identifying data stewards
There are different approaches to identifying individuals for the different roles that have been formalized. Common techniques are as follows:
- The executive sponsor for data governance can send a request to their direct reports explaining the importance of identifying these roles and requesting that individuals be nominated by a certain date.
- Each member of the data governance body can be tasked to identify people for different roles within their area of the organization. A variation of this is for the governing body to identify the Coordinating Data Stewards, and then ask them to identify people for the other roles.
- The data governance lead or deputy can meet with managers of each division, describe the roles, and ask the managers to identify individuals for each role.
See Chapter 7 for examples of different DOT data stewardship domain models
- IT groups may have lists of system owners and/or SMEs that can provide a good starting point.
- Business data stewards and owners/trustees can be identified as part of the process of creating an initial agency data inventory or catalog.
People can play multiple roles. Keep in mind that some individuals may play multiple roles – for example, a single person may be the owner/trustee, business data steward, and custodian for a given data asset.
Some roles may be played by multiple people. Also, there may be situations where two or more individuals play a given role for a given data asset. There may be more than one SME, or multiple IT staff members assigned to a given data system. Some agencies have handled the case of multiple people per role by designating a single steward for a data asset, and instructing them to form a work group or advisory group consisting of the other SMEs as well as custodians. This simplifies the task of tracking (and maintaining) the list of stewards while acknowledging the fact that multiple individuals share responsibility for the data asset.
Keeping your stewards list updated. Given the level of turnover at DOTs due to retirements, employee departures and job changes within the agency, it is likely that your list of stewards will quickly become outdated. Therefore, you will want to always track the name of the business unit for each data steward, and use one (or more) of the following methods to keep the data steward contact information up to date:
- Assign stewardship roles at the business unit level. In this approach, the unit manager serves as the de facto data steward and delegates tasks to members of their staff.
- Assign stewardship roles to specific positions rather than named individuals. Having a specific individual to contact about a dataset is helpful, but if a role within a business unit is stable and it is easy to discover the person in that role, this can be a workable approach.
- Ask each current steward to notify the data governance lead, assistant or Coordinating Data Steward if they are no longer able to serve in their role (and to name a logical successor).
- Send a quarterly or annual email out to each data owner/trustee with a list of the stewards for their data assets, asking them to provide any updates.
- Send a quarterly or annual email directly to each steward asking them to confirm that they are still serving in their position.
- Ask the Coordinating Data Stewards to update the stewards list for their areas (as they are made aware of changes).
- Track employee departures (based on data provided by the Human Resources unit) and follow up with supervisors of any departing employees on the stewards list to identify replacements.
- Build in processes to update the stewards in conjunction with established workflow for maintaining the agency’s data inventory or catalog.
- Coordinate with the IT group on processes to update information system owners or SMEs.
Training and supporting data stewards
Individuals who are identified for data stewardship roles will need to understand what is expected from them. They will also benefit from templates, examples, training and support for meeting these expectations. Specific ways to support your data stewards include:
- Onboarding sessions provided by the data governance staff and/or the Coordinating Data Stewards.
- Fact sheets summarizing expectations for each role.
- Presentations describing how the different roles interact with each other, with examples of specific situations where action from each role is expected.
- Training modules covering overall responsibilities as well as more in-depth topics such as metadata and data quality management.
- Standard tools and templates (e.g., for metadata or business rules).
- Technical assistance from data governance staff with producing any required artifacts or deliverables – can be provided on request or organized as a hands-on training activity where a group of stewards develop required products over several weeks, supported by a technical lead.
- Office hours – designated time for getting support from the data governance staff.
- Creating a Center of Excellence staffed with experts (agency employees or on-call consultants) made available to support data stewards.
Summary
Establishing roles and responsibilities for data governance and stewardship is an essential activity that sets your agency up to achieve the established goals. This chapter has covered the three areas where roles and responsibilities are needed: leadership and staffing to carry out the data governance initiatives, governing bodies that will make decisions and connect data governance to the rest of the organization, and stewardship roles which establish the foundation for improved data management and coordination.
Many agencies find that they need to make adjustments to roles and responsibilities over time:
- Data governance leads may identify the need for new types of skills to support priority activities – or may require additional staff to achieve desired outcomes;
- Data governance bodies may need to add or replace members to get better engagement from a particular part of the organization;
- Data governance bodies may need to be reconstituted to better align with new external requirements or internal shifts in management; and
- Data stewardship roles may need to be adjusted or further clarified following the initial rollout.
You should plan to periodically assess how well things are working and whether changes are warranted.
