data governance articles

A Modern Operating Model for Agile Data Governance

By Kelle O'Neal

Traditional data governance operating models, including centralized, decentralized and hybrids of the two, focus on governance councils and committees. These groups drive the development of policies and processes and oversee enforcement. Today’s needs for agility and speed-of-delivery, paired with greater attention on the value of data, require us to revisit traditional models. We need to ensure their relevancy and utility for data-driven organizations who demand access to self-service data and analytics that reveal insights and opportunities.

The concept of an operating model for governance is still valid: it’s the framework articulating where decisions around data are made and enforced.

Also still appropriate for a data governance operating model:

  • Ensures the right people are represented and creates accountabilities.
  • Articulates decision-making and communication processes for the roles, responsibilities, accountabilities, ownership and decision rights needed to run data governance.
  • Explains the distribution of work and how it gets done.
  • Adapts to meet the organization’s changing needs.
  • Represents the governance framework and isn’t a hierarchical organizational chart.

Agile Data Governance

The operating model we’re seeing today is agile data governance, which enables data democratization and the rise of the data citizen role. This iteration, unlike ones before it, isn’t a top-down approach — it’s bottom-up and addresses data accountabilities as close as possible to the point of data usage and consumption.

It’s a model that enables data creators and data users to participate in governance activities, simultaneously empowering them and expecting their contributions to the corpus of knowledge about the data to improve enterprise-wide data literacy. And they’re encouraged to follow guidelines, rather than rigid, prescriptive procedures.

Agile governance leverages new and improved technologies that monitor and enforce governance practices, while supporting the enterprise-wide demands of different data users and the increasing complexity of data-related concerns.

Re-imagining Governance Responsibilities and Titles

In the shift to modern, agile governance, we still see a need for traditional roles that include data governance leads, data owners and data stewards. But we’re seeing old roles re-imagined (and re-named) and new roles emerging.

For example, in some organizations data stewards are now called data custodians or data guardians. In others, data owners are now known as data champions. These name changes are prompted by greater awareness of the roles and their critical accountabilities, coupled with the desire for more engaging and less authoritarian-sounding titles.

We’re also seeing more granularity in certain roles. A business data steward is equivalent to what’s commonly call a data steward. A technical data steward is the business steward’s IT counterpart who is responsible for tools that support the business data steward. These steward roles work hand in hand to make sure business processes are supported by technology infrastructure, with the technology data steward recommending improvements in processes, based on better and more creative technology usage.

It’s also becoming popular to identify explicit stakeholder roles, such as data consumers, data producers and other participants in data governance, even when they may not be engaged in governance full-time. This is a recognition that their participation is truly important and should be called out.

New, Agile Governance-Inspired Roles

Today, many companies are expanding the C-Suite by hiring a chief analytics officer (CAO) who, as one publication called the role, is the executive who turns data into decisions. (Side note: A quick search on LinkedIn shows 1,689 people with the CAO job title.)

The CAO is usually a counterpart of the chief data officer (CDO). A CAO typically oversees the entire analytics process and environment. The role may report to the CDO, but more commonly we see the role being a peer.

Also new to the governance landscape: a data acquisition lead who is responsible for the end-to-end process of identifying new data the organization needs. The role oversees the negotiation and procurement of the data, the ingestion of the data and ensuring it meets all internal requirements (e.g., governance, legal privacy and security). Another responsibility of the data acquisition lead is making the enterprise aware that data is available through effective communication and by creating a provisioning or access strategy.

The data acquisition lead doesn’t have to just manage data acquired from outside sources either. In many large organizations, this role also pertains to internal data moving around the organization. As data governance has evolved to become a service line to the rest of the business, this role has become increasingly important in facilitating the creation, distribution and delivery of services.

Based on the volume of data needed to support a business, data operations is becoming more important to companies. Data operations can be embedded into analytics groups to speed up the cycle time of data analytics, or it can be focused on data that is used within business operations. In this case, they would be responsible for direct oversight and management of the data, including remediation of data in source systems. Data operations would be responsible for capturing knowledge about data, per the metadata strategy, and responsible for working with peers to resolve cross-functional and cross-divisional data issues and requirements. This may sound similar to the data steward’s role, but the difference is that it’s more hands on — and it’s done as part of a specific process, such as sales or marketing.

Putting New Models and Roles Into Practice

How do we pull together a new structure and evolved (or newly titled) roles to make a data governance organization more modern and more successful?

Here are factors to consider when standing up or reworking the data governance model in your organization:

  • Take into account your company’s existing decision-making processes and structure. This includes organizational complexity and maturity and, possibly, domain complexity and maturity, and how governance can best scale across the organization.
  • Make sure the governance team is cross-functional and includes people from multiple lines of business. This is key to getting buy-in and to ensuring adoption of standards, processes and guidelines you’re creating as a governance organization.
  • Encourage teams meet on a regular basis and hold them accountable for doing so. This ensures a mechanism of a cadence where people regularly come together to identify issues and create and implement solutions, particularly those connected to new and emerging programs so governance remains connected to the business’s real-time needs. This may start to look more like a meet-up than a council meeting and it is important that there is meaningful content to review, decisions to be made and actions to be completed.
  • Promote clear, open communication — and not just with your governance counterparts and first-level stakeholders, but with executives, as well. Also, recognize there are other stakeholders, project managers and potentially other domain stewards who could benefit from some sort of involvement and, at a minimum, at least periodic status updates. Ensure the communication is frequently face-to-face and leverages creative mechanisms, with email being the least frequent.
  • Executive sponsorship is a definite must — they’re the cheerleader, evangelist and key champion of your data governance program and its capabilities. While this can be an individual, the role can also be done through a leadership forum.
  • Regardless of where you are — establishing an operating model or revising an old one — leverage governance work that’s already been done at your organization. This helps reduce the amount of (and the perception of) change the organization requires and makes adoption of a new, improved model much easier.

Modern, agile data governance doesn’t mean there isn’t a need for rules about data — a la a centralized, decentralized or hybrid model approach. Agile governance incorporates the best parts of the established models but shifts the focus to providing more support that empowers individuals to collaborate more and get more value from data.

This is a model whose time has come. We’re seeing its adoption in some of our client organizations and expect even more to follow suit in the coming years.

You have Successfully Subscribed!