Who Owns Master Data Now

Why Master Data Management Has Become an Executive Accountability Issue

Master data used to be treated as an information technology problem. The customer record sat in one system, the supplier record in another, product data lived elsewhere, and finance maintained its own carefully protected version of the truth. When inconsistencies arose, the problem was usually pushed to the technology teams: clean the data, integrate the systems, fix the reports, reconcile the duplicates, and move on.

That view is no longer sufficient.

The latest thinking around master data management points to a far more important conclusion: the organisation that owns its master data owns its ability to decide, act, report, automate and scale with confidence. The organisation that does not own it may have systems, dashboards and artificial intelligence ambitions, but it does not yet have a dependable decision foundation.

Gartner defines master data management as a “technology-enabled business discipline” in which “business and IT work together” to ensure the accuracy, stewardship, governance, semantic consistency and accountability of shared enterprise master data. That definition matters because it does not describe master data management as a software project. It describes it as a business discipline.

This is the central shift executive teams now need to internalise: ownership of master data cannot sit only with information technology. Nor can it sit vaguely with “the business” as an abstract collective. It must be designed as an operating model, with named accountability, defined decision rights, clear stewardship, measurable data quality and executive sponsorship.

For Emergent Africa, this issue sits directly at the intersection of Decision Intelligence, Master Data Management, ESG reporting, strategy execution and digital transformation, all of which are identified as priority themes in the SEO specification for the firm’s thought-leadership and commercial positioning.

The Question Has Changed

The old question was: who owns the system?

The better question is: who owns the meaning?

A customer is not just a name in a customer relationship management platform. A supplier is not just a vendor number in an enterprise resource planning system. A product is not just a stock-keeping unit. An asset is not just a line item. A location is not just a site code. These entities carry commercial, operational, financial, sustainability, risk and strategic meaning.

When the sales team, finance team, procurement team, operations team and sustainability team define the same entity differently, the organisation does not have a technology problem alone. It has a meaning problem. And once you can see that clearly, the ownership question becomes sharper.

Who defines what a customer is?
Who decides when two records refer to the same entity?
Who approves changes to critical supplier attributes?
Who owns the hierarchy used in management reporting?
Who decides which product classifications are acceptable for ESG reporting?
Who is accountable when artificial intelligence uses master data to make recommendations?

These questions cannot be answered by a platform. They must be answered by leaders.

Nicola Askham, widely known as The Data Governance Coach, makes the point plainly in her work on data governance roles: “senior business users” are the true data owners, not information technology teams. Her reasoning is practical. Data owners need enough authority to approve changes, allocate resources and take responsibility for quality.

This does not diminish the role of information technology. It clarifies it. Information technology enables, protects, integrates, automates and scales the master data capability. But the business must own the meaning, policy choices, acceptable quality thresholds and commercial consequences.

Why This Has Become Urgent

Master data ownership has become urgent because the enterprise context has changed.

First, organisations are increasingly multi-platform. Core processes now run across enterprise resource planning, customer relationship management, procurement, human capital, e-commerce, analytics, ESG, data lakehouse and artificial intelligence platforms. Master data moves through all of them. If each platform becomes the de facto owner of its own version, the enterprise slowly loses control of its shared business language.

Second, artificial intelligence has raised the stakes. McKinsey’s 2025 State of AI research found that while artificial intelligence use is broadening, many organisations are still struggling to scale enterprise-level value. The research notes that high-performing organisations are more likely to have senior leaders who demonstrate “ownership of and commitment” to artificial intelligence initiatives. The implication for master data is direct: artificial intelligence value depends not only on models, but on governed, trusted and contextual data foundations.

Third, regulatory and reporting pressure is increasing. ESG reporting, audit readiness, financial controls, privacy obligations, supplier due diligence and operational risk management all rely on trusted master data. If the organisation cannot explain where data came from, who approved it, how it changed and what version was used, it cannot credibly claim that its reporting is controlled.

Fourth, the market is signalling a strategic revaluation of master data. Salesforce’s acquisition of Informatica was positioned around trusted data, governance, quality, metadata management and master data management as foundations for agentic artificial intelligence. Salesforce explicitly linked Informatica’s MDM, quality controls and policy management to data that is standardised, accurate, consistent and secure. SAP’s 2026 announcement that it intends to acquire Reltio made the same point from another direction: SAP said the acquisition is intended to help customers make SAP and non-SAP enterprise data artificial-intelligence ready. Muhammad Alam of SAP said, “AI cannot reach its full potential when data is fragmented.”

These are not isolated technology transactions. They are market signals. The world’s leading enterprise platforms are repositioning master data as the control layer for trusted artificial intelligence, automation and decision intelligence.

The Myth of One Owner

One reason master data ownership fails is that organisations search for a single owner when they actually need layered accountability.

There is no single owner of master data in a complex enterprise. There is an ownership system.

The board and executive committee own the risk, performance and trust agenda. They do not maintain records, but they are accountable for whether the organisation can make reliable decisions, report credibly and scale artificial intelligence safely.

The Chief Data Officer or Chief Data and Analytics Officer owns the enterprise data strategy, governance model and capability roadmap. This role should not become the dumping ground for every data problem. Its purpose is to orchestrate the data operating model, ensure alignment, enforce standards and make data a managed enterprise asset.

Business domain owners own the meaning and quality of master data within their domains. The head of procurement may own supplier master data. The commercial leader may own customer definitions. The finance leader may own the chart of accounts, cost centres and reporting hierarchies. The operations leader may own asset or location data. These owners make policy decisions, approve standards and resolve trade-offs.

Data stewards manage the day-to-day operational integrity of data. IBM describes data stewardship as practices that help ensure data quality and accessibility, with responsibilities including metadata, reference data, lineage, classification and data quality metrics. IBM also notes that stewardship has become more significant as artificial intelligence consumes and produces large volumes of data.

Information technology owns the platforms, integrations, security architecture, workflows, automation, access controls and technical reliability. It turns business rules into executable capability.

Risk, compliance, audit, finance and sustainability teams provide assurance. They test whether controls are working, whether definitions are fit for reporting and whether data can withstand scrutiny.

This layered model feels more complex than saying “the Chief Data Officer owns data”. But it is far more accurate. It also matches how real organisations work. Decision rights, process accountability, systems capability and assurance cannot all sit in one place.

Business Ownership Does Not Mean Business Administration

There is another mistake to avoid. When organisations hear that master data should be business-owned, they often interpret this as asking busy executives to become administrators. That is not the point.

Business ownership means accountability for meaning, policy, value and consequences.

A business data owner should not spend the day correcting addresses, fixing duplicates or manually checking every field. That is stewardship work, supported by workflows, tooling, rules and automation. The business owner’s role is to answer the questions that only the business can answer.

What does “active customer” mean?
Which supplier attributes are mandatory before onboarding?
What is the approved hierarchy for regional reporting?
What level of product classification is required for margin, sustainability and inventory decisions?
What tolerance level is acceptable for duplicate records?
When should a data defect be escalated because it creates operational, financial or regulatory risk?

Once those choices are explicit, they can be turned into rules. Once they are rules, they can be embedded into workflows. Once they are embedded, the organisation begins to move from heroics and reconciliation to repeatability and trust.

This is where master data management becomes a leadership system rather than a cleansing exercise.

The Rise of Domain Ownership

The data mesh movement has influenced current thinking by arguing that data should be owned closer to the business domains that create and understand it. Zhamak Dehghani’s data mesh principles include “domain-oriented decentralized data ownership” and “data as a product”, supported by federated governance.

This thinking is highly relevant to master data ownership, but it needs careful interpretation.

Domain ownership is useful because it places accountability near the people who understand the business context. Finance understands reporting hierarchies. Procurement understands suppliers. Commercial teams understand customers. Operations understands assets. Sustainability understands ESG data requirements. Domain ownership improves responsiveness and makes data quality part of everyday business activity rather than a distant central function.

But domain ownership without enterprise governance can create new silos. Each domain may optimise its own definitions while weakening enterprise consistency. That is why federated governance is essential. IBM describes data mesh as decentralised by business domain, but also notes that central governance standards remain necessary. TechTarget’s 2026 analysis similarly describes a model where domain teams build and maintain data products while federated governance applies company-wide rules for quality, security and interoperability.

For master data management, the practical answer is not centralised control versus decentralised ownership. It is governed federation.

The enterprise defines the non-negotiables. Domains own the contextual detail. The platform enforces standards. Stewardship keeps the data healthy. Leadership resolves trade-offs.

Master Data as the Enterprise Trust Layer

Malcolm Hawker, former Gartner analyst and Chief Data Officer at Profisee, argues that “MDM is more important now than ever” and that it remains relevant even as approaches such as data mesh mature. He also notes that “multiple versions of the truth exist” across organisations, because sales, finance and other functions may view the business through different legitimate lenses.

This is an important nuance. Master data management should not be reduced to the simplistic promise of one universal truth for every situation. Enterprises often need contextual truth. Sales may define customer activity differently from finance. Procurement may classify a supplier differently for contracting, risk, transformation or spend analytics. Sustainability may need supplier, product and location data at a different level of granularity from operations.

The goal is not to eliminate all contextual variation. The goal is to govern it.

A mature master data capability allows the organisation to know which definition is being used, why it is valid, where it applies, who approved it and how it connects to other views. This is why the phrase “single source of truth” can be misleading unless it is supported by semantic governance.

The better aspiration is this: a governed system of trusted context.

That phrase matters. Artificial intelligence does not need only records. It needs context. ESG reporting does not need only fields. It needs lineage and assurance. Decision intelligence does not need only dashboards. It needs confidence in the definitions beneath the dashboard.

SAP’s recent master data playbook describes master data management as “not an IT checklist” but a strategic capability that must be designed, resourced and measured. It also recommends executive sponsorship from the business to enforce policy, arbitrate trade-offs and sustain funding.

That is the leadership move many organisations still need to make.

The Platform Cannot Be the Owner

One of the most dangerous assumptions in enterprise data is that the system holding the data owns the data.

It does not.

A customer record in Salesforce, a supplier record in SAP, a product record in an e-commerce platform and an employee record in a human capital system may all be system records. But the enterprise owns the meaning. The platform may be authoritative for certain processes, but it should not become the unchallenged authority for enterprise definition.

BCG’s work on digital platform lock-in highlights why this matters. In its 2025 analysis, BCG reported that 77% of firms prefer to retain their own data model rather than adopt a platform-native data model, because portability, data loss, security and integration risks are strategic rather than merely technical.

This is especially important for South African and African enterprises that often operate across hybrid environments, legacy platforms, regional businesses, acquisitions, shared services, multiple enterprise resource planning instances and varied reporting requirements. If master data is owned by whichever system happens to dominate at a point in time, the enterprise will struggle to create a platform-neutral decision foundation.

The platform should execute the rules. It should not quietly become the source of enterprise meaning.

The Practical Ownership Model

Emergent Africa’s view is that organisations need a practical master data ownership model built around seven disciplines.

First, name the executive sponsor. This should be a senior leader with enough authority to link master data to business outcomes: trusted reporting, artificial intelligence readiness, ESG assurance, customer experience, working capital, procurement discipline, operational efficiency or risk reduction.

Second, appoint domain owners for each critical master data domain. Customer, supplier, product, asset, employee, location, chart of accounts and reference data domains may require different owners. These should be senior business leaders, not technical custodians.

Third, define decision rights. The organisation must know who approves definitions, who can request changes, who resolves conflicts, who signs off data quality thresholds and who has authority when one business unit’s preference conflicts with enterprise standards.

Fourth, formalise stewardship. Data stewards need time, tools, workflows and community. They are not volunteers cleaning up the mess after the fact. They are operational guardians of data quality, working with business owners and technology teams to keep master data usable.

Fifth, embed governance into the data lifecycle. Informatica’s master data governance guidance argues that master data governance applies governance directly where finance, procurement, customer operations and analytics intersect. It also distinguishes between governance as rules, master data management as authoritative records and master data governance as the operationalisation of rules on critical data.

Sixth, measure what matters. Data quality should be visible in executive routines. Duplicate rates, completeness, approval cycle times, defect ageing, lineage coverage, data issue resolution and business-impact metrics should be tracked. If master data quality is not measured, it will become invisible until it fails.

Seventh, connect master data to decision intelligence. Master data ownership is not an end in itself. It exists so leaders can see clearly, decide faster, trust reporting, automate safely and execute strategy with fewer hidden frictions.

The Executive Test

Here is a simple executive test.

Ask the leadership team these questions:

Can we identify the official owner of each critical master data domain?
Do our data owners have real decision rights?
Do our stewards have enough capacity to do the work properly?
Can we explain where our most important ESG, finance, customer and supplier data came from?
Do we know which definitions are being used in our key dashboards?
Can artificial intelligence safely use our master data without amplifying inconsistency?
Do we have a controlled process for changing master data definitions?
Can we separate platform ownership from enterprise data ownership?

If the answers are unclear, the organisation does not yet have a mature master data ownership model.

And once leaders see that, the next step becomes obvious: stop asking information technology to carry an enterprise accountability problem alone.

The South African Executive Relevance

For South African executive teams, the issue is particularly relevant. Many organisations are under pressure to improve growth, manage costs, strengthen resilience, meet ESG expectations, modernise technology, use artificial intelligence and improve strategic execution. Each of those priorities depends on trusted data.

A retailer cannot personalise effectively if customer data is fragmented.
A mining company cannot report confidently if asset, supplier and sustainability data are inconsistent.
A healthcare group cannot improve operational performance if patient, provider, facility and supplier data is unreliable.
A financial services group cannot scale artificial intelligence responsibly if its customer and product records are not governed.
A diversified industrial group cannot improve decision quality if every division defines the same entities differently.

Master data ownership is therefore not a back-office debate. It is a boardroom issue hiding inside operating detail.

When leaders begin to treat master data as an enterprise capability, something shifts. Reporting becomes less contested. Decision meetings become less defensive. Transformation teams spend less time reconciling and more time improving. Artificial intelligence initiatives become more credible. ESG claims become more auditable. Strategy execution becomes less dependent on heroic manual correction.

That is the point of ownership. It creates confidence.

The Emergent Africa Point of View

The next phase of master data management will belong to organisations that can combine central governance with business ownership, technology enablement with executive accountability, and data quality with decision intelligence.

Emergent Africa’s position is that master data management should be treated as a managed enterprise capability, not a once-off implementation. The capability must be platform-neutral, business-led, governance-enabled and connected to measurable outcomes. This is particularly important in multi-platform environments where enterprise data must move across operational systems, reporting platforms, sustainability tools, analytics environments and artificial intelligence use cases.

The organisations that move first will gain more than cleaner records. They will gain a trusted foundation for faster decisions, stronger governance, better reporting, improved automation and more disciplined execution.

The ownership question is therefore not: who owns the database?

The real question is: who owns the organisation’s ability to trust what it knows?

When that question lands properly, master data management stops being a technical clean-up exercise and becomes a leadership imperative.

Conclusion

Master data ownership is entering a new era. The business must own meaning. Technology must enable control. Stewards must protect quality. Governance must define the rules. Executives must sponsor the operating model. Boards must recognise the link between trusted data and trusted decisions.

As artificial intelligence, ESG reporting, digital transformation and strategy execution all converge on the same data foundation, organisations can no longer afford blurred accountability. The companies that win will not be those with the most systems. They will be those with the clearest ownership of the data that those systems depend on.

Emergent Africa helps organisations strengthen the connection between master data management, decision intelligence, governance and strategic execution. To explore how a clearer master data ownership model could improve decision quality, reporting confidence and artificial intelligence readiness in your organisation, connect with Emergent Africa.