Driving Operational Clarity with Effective Enterprise Architecture and Data Governance

Data governance is more than just day-to-day monitoring; it is about ensuring ongoing integrity, accessibility, continuity and security.

Data governance is more than just mapping enterprise processes across platforms, data models and structures; it is about strategic alignment, master data integrity, accessibility, continuity and security, process improvement, and cross-functional and cross-silo integration. (Stock image.)

Top-floor to shop-floor data analytics is on everyone’s mind when it comes to data-driven decision making and effective operations. There is, however, a lot more to enterprise architecture and data governance when it comes to optimizing operations. Ensuring that customer, supply chain and product data is fit for purpose and is being used properly is core to data stewardship responsibilities. This includes how data is structured, how data and business models are configured and customized, and how data is authored and consumed across operational processes—as well as across digital platforms and interfaces.

In this article, I elaborate on enterprise architecture and its associated data governance, what it means to engineering and manufacturing organizations and what value it might bring to product development operations.

Data governance is part of enterprise architecture—not to be confused with data architecture and data governance, which are typically defined as follows:

• Enterprise architecture refers to all operational aspects of managing a business, from leadership to related processes; and data and tools, from vision to requirements and standards. This is quite a broad definition, which clearly encompasses platforms and tools but also goes far beyond IT as a support function.
• Data architecture refers to enterprise data states and their associated transformation rules. It describes how data flows across functions, translating strategies and requirements into technical elements that connect data across platforms. These elements include hardware, storage, computing, hosting, networks, infrastructure, software, apps, access, usage processes, visualization, interaction, an enterprise service bus and many more integration requirements.
• Data governance refers to how data ownership, standards and policies inform how enterprise data will be structured, accessed, created, consumed, shared, integrated, maintained, archived, and so on.

Having a clear enterprise architecture strategy can be a significant differentiator to help ensure that business strategies and operations align toward effective data implementation. Master (or reference) data strategies enable organizations to rapidly evolve and take advantage of agile processes and technologies, seamlessly aligning IT systems and business models for ongoing change leadership and transformation.

The Enterprise Architecture Book of Knowledge (EABOK) consortium, which ceased operations in October 2020, simply defined enterprise architecture as “an organizational model [in the form of] an abstract representation of an enterprise that aligns strategy, operations and technology to create a roadmap for success; . . . an enterprise being a complex organization that is attempting to undergo change.” This is to help organizations “leverage the full power of the data they already have available” and “bring together the enterprise knowledge (vested in its people, policies and operations), with knowledge of technology to improve the business and enable key decision makers to effectively steer the organization.”

Additionally, Gartner defines enterprise architecture as “a discipline for proactively and holistically leading enterprise responses to disruptive forces by identifying and analyzing the execution of change toward desired business vision and outcomes . . . delivering value by presenting business and IT leaders with signature-ready recommendations for adjusting policies and projects to achieve targeted business outcomes that capitalize on relevant business disruptions.”

Although the Gartner definition may not be the most elegant one, it does highlight the need for standards, policies and frameworks to manage change, both internally and externally sourced, so that organizations can anticipate and adapt to disruptive forces. More importantly, enterprise architecture is about clear data, processes and platform ownership, and translates into a communication, decision-making and continuous improvement framework.

Some enterprise architecture frameworks can be adapted to a given organizational context:

• The architecture development method (ADM) of The Open Group Architecture Framework (TOGAF)  introduces the vision, business architecture, data and application architecture, and architecture technology. It provides a method for developing and managing the life cycle of enterprise architecture.
• The Zachman Framework provides a template matrix to align roles/persona perspectives with enterprise integration elements. It includes key high-level questions (what: data; how: function; where: network; who: people; when: time; and why: motivation) to help manage and organize key business relationships—from scope to operating models, information systems, technology models and detailed data representations.
• The 4+1 view model of software-intensive systems links multiple architecture levels to reflect and align different stakeholder perspectives, from the logical view (end users, customers and data SMEs) to the process (functions and operational performance), physical (platform mapping to hardware) and development (system implementation) views.

It is often said that modern data architecture equals cloud-based platforms and data lakes, which can be adapted to support data growth and processing requirements. Modern platforms aim to deliver on-demand storage scalability (vertical scaling), as well as the ability to expand to support increasing processing requirements (horizontal scaling).

Understanding and optimizing how data and metadata are mapped across multiple sources provide clarity on how data is gathered, processed and integrated across enterprise platforms. A robust data architecture focuses on data classification and transformation rules across “single sources of data” and their associated platforms.

In one of its recent articles, McKinsey (2020) highlighted six fundamental shifts that can help organizations create effective data architecture strategies, helping them transition from:

1. On-premise to cloud-based platforms
2. Asynchronous to synchronous/real-time data processing
3. Pre-integrated commercial solutions to modular, best-of-breed platforms
4. Point-to-point to decoupled data access
5. An enterprise warehouse to a domain-based architecture
6. Rigid data models toward flexible, extensible data schemas

Furthermore, the McKinsey report suggests that implementation speed is critical to “rapidly evaluate and deploy new technologies so that [organizations] can quickly adapt” to new data paradigms by:

• Applying a “test-and-learn mindset”
• Establishing “data tribes with effective data stewards, data engineers, and data modelers”
• Investing in “DataOps to accelerate design-development-deployment of new components of the data architecture”
• Creating a “data culture” to encourage the use of new data services within day-to-day operations with enhanced analytics and embedded intelligence

When it comes to running operations, data governance contributes to continuous data-driven improvements, linking business value to data quality and enforcing the relevant data metrics and remediation actions. It is also about continuously adjusting the target operating model and its associated data policies to reflect ongoing changes.

Boston Consulting Group (BCG) describes data governance as four building blocks: 

1. Data structures and domains
2. Data policies
3. Data tools
4. Data organization stakeholders and the target operating model (TOM)

“Data governance may include organizational and technological elements that facilitate the sustainable improvement of a company’s data quality” (BCG, 2019).

Effective data governance aims to reduce errors and improve process and data quality compliance. As highlighted by the BCG report mentioned above, data governance goes beyond policing data usage; it also includes the optimization of associated “processes, actions, roles, and budget allocation principles.”

What are your thoughts?

• Kruchten, P. (1995). Architectural Blueprints: The 4+1 View Model of Software Architecture.
• The Open Group. (2001). Introduction to the Architecture Development Method (ADM).
• Zachman, J. (2008). The Concise Definition of the Zachman Framework.
• McKinsey. (2020). How to Build a Data Architecture to Drive Innovation—Today and Tomorrow.
• Boston Consulting Group. (2019). Good Data Starts with Great Governance.