Written by Bob Kaplan and Suchitra Deo
In our previous article, we argued that AI success is not driven by better models, but by better data. The natural next question for leadership teams is: what does it actually take to operationalize that insight?
Recognizing that data matters is straightforward. Building the system to manage it at scale is where most organizations struggle.
Many companies have already invested heavily in data initiatives, yet the outcomes often fall short. The root cause is not a lack of effort, but a lack of an effective operating model. Ownership of data is often unclear or diffused across functions. Accountability exists in theory but is rarely enforced in practice. Data issues are addressed downstream, after they have already created friction, rather than being resolved at the source. Governance bodies convene, but decisions do not consistently translate into changes in workflows or systems.
As a result, organizations see incremental improvements, but not the performance gains required to support AI and advanced analytics at scale.
Making governance operational
Effective data governance is often misunderstood as a set of policies or committees. In reality, it is a system of decision rights and accountability that must function day-to-day within the business.
Business ownership:
The business (not just IT) must define what matters. This includes establishing standard definitions for critical data such as customers, products, and revenue, determining how that data should be used in decision-making, and setting expectations for quality and consistency. Without this, different parts of the organization will continue to operate on competing versions of the truth.
Data stewardship:
This is the operational layer that ensures standards are applied in practice. Stewards resolve inconsistencies, manage exceptions, and validate updates as data moves through systems. Importantly, this work happens in real time, embedded in workflows, not as a periodic cleanup exercise.
Central governance layer:
This function ensures consistency across business units, resolves conflicts when definitions or priorities diverge, and focuses data efforts on areas that drive the most value. Without this coordinating mechanism, organizations tend to drift back into fragmentation as they grow.
When these 3 elements work together, governance moves from being a passive oversight function to an active operating capability.
Master data management: turning governance into reality
While governance defines expectations, it does not change the data itself; that is the role of master data management (MDM). At its core, it establishes a single, authoritative version of key business entities (customers, products, accounts) across the organization. In most companies, these entities are duplicated and inconsistent across systems, from CRM and billing platforms to operational and analytics tools.
MDM addresses this fragmentation by creating a “golden record,” a unified version of each entity that is continuously reconciled across sources. It applies defined business rules to determine how conflicts are resolved, which data sources are authoritative, and how updates are validated. Critically, it does this upstream, preventing issues from propagating rather than fixing them after the fact.
A well-implemented MDM capability acts as a control layer across enterprise systems. It unifies fragmented data, eliminates duplication through automated matching and merging, and synchronizes clean, consistent data back into operational platforms so that every team is working from the same information.
The impact is immediate and tangible. Teams no longer spend time debating which system is correct. Sales, operations, and finance align around a shared view of the business. Systems like Salesforce and analytics platforms begin to deliver their full value because they are no longer constrained by inconsistent inputs.
Why this matters for business leaders
MDM is not just a technical initiative, it is a core business capability with direct impact on performance.
When data is not governed and unified, the consequences are visible across the organization. Commercial teams operate on incomplete or conflicting customer information. Employees lose time to manual reconciliation and rework. Compliance and reporting processes become more complex and costly. Technology investments, from CRM to AI, fail to deliver their expected return because they are built on unreliable inputs.
When governance and MDM are implemented effectively, the impact is structural. Teams operate with a consistent, trusted view of the business. Decision-making accelerates because validation cycles are reduced. Systems function as intended, enabling rather than constraining performance. Most importantly, AI initiatives can scale with confidence, because their outputs are grounded in reliable data.
This is not a marginal improvement – it is a shift in the organization’s ability to execute.
Building the foundation for execution
At HighPoint Associates, we work with leadership teams to move beyond data strategy and into execution. We help organizations establish clear ownership and governance structures, design and implement MDM capabilities aligned to business priorities, and embed data management into core workflows and systems.
This is ultimately an operating model shift. Organizations that get it right reduce time spent reconciling information, increase alignment across teams, and perhaps most importantly, unlock the full value of their technology and AI investments.
Because in the AI era, competitive advantage will not come from recognizing the importance of data. It will come from the ability to operationalize it – consistently, at scale, and in a way that drives real business outcomes.
