Data Capability Is Not Data Technology: Lessons from IKEA’s Global Data Transformation
Q1. IKEA is one of the world’s most recognizable global brands — present in dozens of countries, with an enormously complex supply chain, a vast network of product suppliers, and data distributed across multiple systems and geographies. When you took on the role of Global Data Development Manager, what was the most surprising thing you discovered about the actual state of data capabilities at that scale — not the version that appears in a transformation roadmap, but the messy reality on the ground — and what did it tell you about what genuinely needs to come first before any data management framework can take hold?
Naveen Gupta: What surprised me was not the complexity itself. At IKEA’s scale, complexity is inevitable. What surprised me was how differently the same data could be understood, owned and managed across different parts of the organisation.
Transformation roadmaps tend to make data management look structured: domains, governance, platforms, data products and target architectures. The reality is more complicated. Data sits across legacy and modern platforms, definitions vary, dependencies are not always visible, and accountability and maturity differ across the organisation.
That taught me that transformation should not begin with technology. It should begin with business meaning and accountability.
Before asking which platform we need, we should ask: What are our critical data objects? What do they mean? Who owns them? What does good quality mean for the people using them? And who has the authority to resolve disagreements?
My biggest lesson is simple: start with accountability, business meaning and real business problems. Then build the framework around them. Otherwise, governance risks becoming documentation rather than actually governing anything.
Q2. You have been driving the implementation of a global data management framework, data quality, data governance, and data enablement capabilities including catalogue, glossary, lineage, and quality — simultaneously. That is an enormous scope. For a data leader at a large organization who is trying to decide where to start and what to sequence, what is the most important lesson you have learned about the order in which these capabilities need to be built — and what happens to organizations that try to build the analytical layer before the foundational data management layer is stable?
Naveen Gupta: There is always pressure to start where value is most visible: dashboards, advanced analytics and now AI. But those capabilities are ultimately consumers of everything underneath them.
If master data is inconsistent, definitions are unclear, ownership is missing and quality is not measurable, sophisticated analytics can simply become a sophisticated way of presenting the inconsistency.
I see the sequence as ownership and governance first, then standards and critical data, data quality, metadata and lineage, integration and finally scalable analytics and AI.
That does not mean waiting years before delivering business value. Foundations and use cases should progress together. But the dependencies need to be understood.
One of the most damaging patterns is repeatedly cleaning poor data downstream. Every workaround creates another transformation, another reconciliation and potentially another interpretation of the truth. Eventually users lose trust and create spreadsheets and alternative reports.
My rule is: move quickly toward value, but never so quickly that you industrialise ambiguity.
Q3. You describe your role as including the task of spearheading data culture and data literacy across IKEA — which is a very different challenge from implementing a technology framework. Culture change in a large, globally distributed organization is notoriously slow and difficult, and data literacy programs have a mixed track record in enterprise environments. What has actually worked at IKEA to move the needle on data culture — not the initiatives that looked good in a presentation, but the ones that changed how people actually behave with data in their daily work — and what would you do differently if you were starting that journey again?
Naveen Gupta: The biggest lesson is that data culture does not change because people complete data training. It changes when their everyday behaviour with data changes.
We have found that leadership visibility, clear business ownership, practical data discussions and real examples are much more powerful than simply explaining a governance framework.
I often start with a basic question: “What is in it for me?”
Someone working in finance, procurement, logistics or product development needs to see how better data makes their work easier, faster or more reliable. When data quality prevents an operational issue, reduces manual correction or improves a decision, the value of data management becomes tangible.
I also believe data quality has to become a collective responsibility. Everyone who creates, changes or consumes data influences its quality.
If I were starting the journey again, I would spend even less time initially explaining the complete framework. I would identify a handful of painful business problems, solve them with the people experiencing them, make ownership visible and use those successes to demonstrate the framework.
Culture changes when data management becomes useful, not simply when it becomes mandatory.
Q4. You are responsible for the Data Stream implementation for S/4 HANA — covering SAP MDG, SAP Datasphere, and SAP Analytics Cloud — across data management, analytics, and integration. SAP transformations at global scale are among the most complex and high-risk technology programs any organization can undertake, and data migration and data quality issues are consistently the primary cause of S/4 HANA implementation delays and cost overruns. What are the two or three data-related decisions that organizations consistently get wrong in an S/4 HANA program — the ones that seem reasonable at the time but create the most painful consequences later — and how have you approached them at IKEA?
Naveen Gupta: I would highlight three.
The first is treating data migration as a project phase rather than a transformation discipline. Profiling, ownership, cleansing and mapping need to start early. Moving existing data and deciding what future master data should look like are two very different activities.
The second is underestimating master-data and enterprise-structure harmonisation. Small differences in identifiers, definitions or organisational structures can create mappings, transformations and validations throughout the landscape. You begin paying what I call a “mapping tax” everywhere.
The third is assuming technology standardisation automatically creates data standardisation. Two SAP environments can still represent the same business differently. That creates complexity across governance, integration and analytics.
My approach has therefore been to treat Data Management, Analytics and Integration as interconnected rather than independent technical workstreams.
Technologies such as SAP MDG, SAP Datasphere and SAP Analytics Cloud are powerful enablers, but they cannot substitute for common definitions, quality and accountability.
Data is not a by-product of an S/4 transformation. It is one of the prerequisites for making the transformation work.
Q5. You manage product suppliers including Microsoft and SAP, run RFPs for product and service procurements, and handle budget, stakeholder, risk, resource, and integration management alongside the technical data leadership responsibilities. That combination — being simultaneously a data strategy leader and a procurement and vendor management executive — is unusual and demanding. What is the most important thing you have learned about managing the relationship between a global organization and its most strategic data technology vendors — and where do you see the biggest gap between what those vendors promise and what they actually deliver in a complex, multi-country enterprise deployment?
Naveen Gupta: The most important lesson is that a strategic vendor relationship should never mean outsourcing your strategic thinking to the vendor.
Major technology providers and implementation partners bring enormous expertise. Their product roadmaps, reference architectures and accelerators can create significant value. But their perspective naturally begins with what their technology can do.
Our responsibility begins somewhere else: with the business problem, our existing architecture, our commercial environment and the capabilities we want to retain internally.
The biggest gap between vendor promises and enterprise reality is often the difference between product capability and operational capability.
“The product supports it” is only the first question. In a global enterprise you also need to ask: Does it fit our data model? Does it integrate with our landscape? Who operates it? How is it governed? How does licensing work across entities and countries? What skills must we maintain internally? And what is the long-term cost of change?
My principle is therefore: treat vendors as strategic partners, but retain architectural judgement, data ownership and commercial intelligence inside your own organisation.
A strategic partnership works best when both sides are strong.
Q6. You are a Data Evangelist, keynote speaker, and content creator alongside your full-time leadership role at IKEA — and you describe yourself as passionate about sharing the lessons of data transformation with a broader audience. Looking back at nearly four years in this role, what is the single most counterintuitive insight you have gained about what it actually takes to build genuine data capability inside a large global organization — the insight that contradicts what most frameworks, consultants, and conference presentations say, but that you have found to be consistently true in the real work of making data valuable at scale?
Naveen Gupta: My most counterintuitive lesson is that building data capability is much less about data technology than most data frameworks suggest.
Technology, architecture, governance and quality tools all matter. But sustainable capability comes from people understanding why data matters to their work and accepting accountability for it.
We often measure maturity through implementation milestones: catalogue deployed, lineage implemented, quality platform established, governance framework published. Those are useful milestones, but they are not the outcome.
The real test is what happens when the central data team is not in the room.
Does someone challenge an inconsistent definition? Does an owner investigate a quality problem instead of working around it? Does a project consider master data before designing an interface? Does a manager question a KPI whose origin is unclear?
If those behaviours happen, you have built data capability. If they do not, you may simply have built data technology.
After working across governance, analytics, master data and large-scale transformation, I am therefore less interested in whether an organisation can show me its data framework.
I am more interested in whether people can consistently answer three questions:
What does this data mean? Who is accountable for it? And do we trust it enough to make a decision?
If an organisation can answer those questions, technology can accelerate it enormously. If it cannot, technology often just accelerates the confusion.
……………………………………..

Naveen Gupta is a global data and digital transformation leader currently serving as the Global Data Development Manager and Global Data Lead at the Inter IKEA Group in Basel, Switzerland. With over 15 years of international experience, he specializes in enterprise data strategy, data governance, SAP S/4HANA data migrations, and building data-driven organizational cultures.