Data Governance, Data Leadership or Data Architecture: What Matters Most?

By on

Robert Seiner and Anthony Algmin faced off – in a virtual sense – at the DATAVERSITY® Enterprise Data World Conference to determine which is more important: Data Governance, Data Leadership, or Data Architecture. Initially planning to play the theme from “Rocky” as they emerged from backstage wearing capes, these two heavy hitters instead battled it out online with each having four rounds to make a case.

Seiner is President and Principal of KIK (Knowledge is King) Consulting & Educational Services and the publisher of The Data Administration Newsletter (TDAN). Seiner is an author, educator, and thought-leader in the fields of Data Governance and Metadata Management and proponent of Non-Invasive Data Governance.

Algmin is the Convergence Platform Program Lead for AbbVie, a Data Leadership advocate, author of the first published book on Data Leadership, and host of the Data Leadership Lessons podcast. Serving as a consultant, data architect, and Chief Data Officer, Algmin has led data-driven transformations in many industries.

By way of an overview, Seiner put forth these questions:

  • Why tie Data Governance, Data Leadership, and Data Architecture together?
  • Is it possible to have one of these things without having the other two?
  • Why is it important to have a succinct, easily understandable definition?
  • What is the relationship of each one to the others?
  • Which of them is the most important?

Round One: Definitions

Data Governance

Even if your organization has no formal Data Governance, data is still being governed. Decisions made by default still affect the business, but not necessarily in productive or profitable ways. Governance provides a foundation, creating rules and structure for data use. “Data Governance is the execution and enforcement of authority over the management of data and data related assets,” said Seiner. The strong wording is important, he said; no matter what approach you take, “You need to execute and enforce authority over the management of data.” It’s not an area where an “agree to disagree” attitude is effective. Without the execution and enforcement, “You won’t solve anything.”

Algmin agreed and added that in the enforcement of authority, Governance provides the “what”: “What are we prioritizing? What’s important about our data?” Both agreed that a succinct, easy-to-understand definition is critical, because, as Algmin said, “This stuff is complicated.” The collective actions of the people working with the data drive the business processes that lead to success, and they need clear definitions to properly carry out those processes.

Data Leadership

Seiner said that Data Leadership is what’s needed from management – people who have a plan and know how to go about implementing that plan. Algmin said that Data Leadership is really about maximizing data value. “Data value” is the difference in outcomes that a business achieves by using data compared to what they would achieve if they didn’t use that data. That value is reflected in increased revenue, decreased costs, or improved risk management. “So, Data Leadership is the ‘why’ are we doing this: we’re creating data value.” In some organizations, Data Leadership is entirely missing, he said.   

Data Architecture

Data Architecture often consists of models and policies and rules and standards that are used to govern data. In order to successfully govern which data is collected, how it’s being stored, how it’s being arranged, how it’s being integrated, how it’s put to use in your systems, “Data Architecture is critical,” Seiner said.

Algmin added that Data Architecture is where key artifacts and mechanisms are used to make Data Governance

not just happen once, but happen in a consistent, reliable, and extensible way to benefit the business. “These are are core foundational business capabilities that drive real meaningful outcomes,” he said. It’s the ‘how.’”

Round One Summary

All three are required by the organization, Seiner said. It’s important to put governance in place so that people behave appropriately around the definition, production, and usage of the data. Data leadership that understands what is being done and why it’s important, with a plan for how to achieve it, is important as well. With a formal governance program, leadership that understands how to manage data, and a formal method for implementing Data Architecture, “You’ve really hit the trifecta, and that’s where we’re really going to be successful” Seiner said.

They’re all parts of the same story, added Algmin. Although Algmin believes leadership is the most important of the three, leadership isn’t effective without Data Governance and Data Architecture.

Round Two: How Leadership Complements Data Governance and Data Architecture

When Seiner defines best practices around Data Governance, the number one success indicator he’s found with every client is that senior leadership supports, sponsors, and understands what what’s involved in implementing Data Governance or Data Architecture. Oftentimes organizations will not be given the time or the resources necessary to implement a Data Governance program or a Data Architecture unless the value is understood at the highest levels of the organization, he said.“ So we, as practitioners, understand that we need to communicate effectively with our leadership as to what these disciplines are, and how they relate to each other.”

Data Governance includes the enforcement of authority. Data Architecture starts to formalize it, and put the right pieces in place, so that it can happen consistently, said Algmin. Company leadership—not data leadership—should decide what’s important for the organization as a whole, but don’t expect them to know the best way to use data to achieve those goals. Data Leadership must find ways to create a story that combines Data Governance, Data Leadership, and Data Architecture and provides solutions to the problems company leaders already know they have, he said. It starts with asking how to better achieve the mission of the business, and finding where data plays a role in that mission.

Business leaders need to know what decisions they need to make and what investments are required to make the most of the data. If a business leader is struggling to meet service level agreements with their clients, “We can come up with a solution to that,” said Algmin. Leadership doesn’t need to learn all the functional differences and details of governance and architecture, they just need viable solutions that improve the business. Ultimately, the most meaningful change results from effectively maximizing the potential from one step of the data lifecycle to another.

Seiner and Algmin agree that without all three disciplines working in a coordinated fashion, it’s impossible to maximize the value of the data. There are some organizations that have done one without the other, but not as effectively if they’d as if they did them all, or did all three of them at the same level. Seiner said, “We want people at the leadership level of our organization to understand how they fit together, why they fit together, and why they’re all necessary,” and that resources are necessary in order to be successful in implementing any of these disciplines across the organization.

Round Three: Incorporating Leadership and DA into DG strategy

To kick off round three, Seiner introduced seven core components of a Data Strategy:

  1. Define business requirements and strategic goals for managing data effectively across the organization.
  2. Understand what questions the business is asking and which of those questions can be answered with data. “Everybody’s heard the phrase, ‘data is the most valuable asset to the organization,’” yet without an understanding of what the business is trying to get out of the data, it’s like shooting at a target that hasn’t been defined yet, Seiner said.
  3. Determine technical or technology infrastructure requirements. Put the technology infrastructure in place to be able to address the requirements that are defined.
  4. Learn how to turn the data into insights, and insights into wisdom.
  5. Determine which people, processes, skill sets, and resources are necessary, and incorporate those into the data strategy.
  6. Set a course for how people will define, produce, and use data across the organization, relying on Data Governance as a meaningful part of the overall Data Strategy.
  7. Use Data Strategy as a roadmap for meeting company goals, incorporating governance, leadership, and architecture with available resources.

Algmin believes that there is a fundamental misunderstanding about the meaning of Data Strategy and argued that it doesn’t exist. Instead, he sees it as a means to an end, because by itself a Data Strategy creates no actual value unless it’s well anchored to a business strategy. “So in the end, it’s really not a ‘data’ strategy at all. It’s a data execution plan.” The end state is already largely defined by the business strategy and determining the execution is all that’s left. “It’s how we take part in that journey that really matters.” Instead, he suggested devising a strategy to measure how the use of data impacts desired company outcomes and contributes to reaching company goals.

Round Four: And the Winner Is…

So which is the most important: Data Governance, Data Leadership, or Data Architecture? While Algmin leans toward leadership and Seiner leans toward governance, both wholeheartedly embrace the metaphor of the three-legged stool and agree that striking a balance among all three is the answer. And although not everyone needs to be a data scientist or a Data Governance professional, said Algmin, “collectively we need to cover all of that ground.”

Seiner agreed, adding that success depends upon competent people dedicated to taking responsibility for each of these areas. “The data will not govern itself, the metadata will not govern itself: You need people to get engaged in doing these things,” said Seiner, and they are all very closely related to each other.

Algmin recommended thinking about which goals are most important—as part of an organization and as an individual—and finding ways to move forward on those a little bit each day, each hour, or in each meeting.

“When can do more of that, it starts to create real momentum.”

Want to learn more about DATAVERSITY’s upcoming events? Check out our current lineup of online and face-to-face conferences here.

Here is the video of the Enterprise Data World Presentation:

Image used under license from

Leave a Reply