As enterprises move from centralized data platforms to domain-driven architectures, data products are becoming the backbone of scalable analytics and AI. The shift is not just technical. It requires rethinking ownership, governance, and how data is created, managed, and consumed across distributed teams.

At Henkel, this transition plays out within a global organization spanning industrial and consumer businesses, where adhesive technologies and consumer brands operate across diverse markets. With more than a century of operational scale, the company must balance innovation with consistency, ensuring that data products support speed, reliability, and compliance across geographies.

In the second installment of this three-part interview series, Katrin Botzen, Corporate Director, Global Data and Analytics at Henkel, in conversation with Julian Schirmer of OAO, examines how the company is operationalizing data products in practice.

Building on the first conversation’s focus on why data and AI matter, Botzen shifts to execution: cloud foundations, product-oriented data structures, governance models, and the cultural work required to keep a decentralized ecosystem aligned as analytics moves closer to where data is created and used.

A cloud foundation that changes what “analytics” means

Botzen starts by placing Henkel’s transformation on a timeline. She says the organization has been fully in the cloud with all applications since last summer, adding that the data and analytics organization has been cloud-based for more than eight years.

That shift changes what analytics looks like compared to eight or ten years ago. Back then, she says, companies were “lucky” when they had only one business data warehouse supported by reporting and “maybe a bit of machine learning on top.”

Now the picture is fundamentally different. “You have basically analytics everywhere where you go,” Botzen says. Analytics is no longer a single entity housed in a single platform. It increasingly lives “closest to the data,” inside operational systems such as CRM, supply chains, and HR.

For Henkel, that decentralization creates both risk and upside. She frames the challenge as governing the “quality ecosystem” while enabling data accessibility for analytics and AI use cases, all while maintaining security and compliance. Done well, Botzen adds, it becomes “a huge opportunity” because it allows the business to solve key problems with analytics and AI.

Data products as shared assets, not one team’s output

Speaking of data products, Botzen is quick to attach a practical label to the term: “Data products are assets currently living not only in one place.”

She notes that Henkel is comparatively consolidated. Botzen points to “one business data warehouse,” Power BI reporting, and a data platform with “a lakehouse concept underneath,” which she describes as an advantage compared to landscapes where “multiple hundreds of data lakes” exist in isolation.

Within that environment, Henkel works with a data mesh concept, designed to align data products with domains and clearly assign responsibility. In the best case, Botzen explains, data products are aligned with domains and “there is a team, looking after exactly this data product,” ensuring it is updated, owned, and ready to be used for analytics and AI.

But the model only works with discipline. Botzen offers a warning in plain language: “Don’t mess up with the mesh.”

Governance through contracts and transparency

In Botzen’s operating model, governance is not a separate layer applied at the end. It is built into the way data products are created, updated, and shared. Henkel uses data contracts to formalize expectations, including refresh cadences and ownership.

The objective is to prevent complexity from turning into opacity, especially as teams combine data products or build new semantic layers. Without clear rules, Botzen notes, it becomes easy to lose track of what exists and who is responsible.

That is why transparency becomes a strategic priority. She points to a data catalog on the core data platform, alongside a broader initiative currently underway. Her aim is consistent: “No matter where your data is, you see where it comes from, who the owner is, what’s in there, the KPIs, and the definition.”

For Botzen, this is not documentation for documentation’s sake: “That’s the ultimate goal, to create that transparency in order to discover what’s there in order to use, and create value.”

Shared standards in a global company

Schirmer introduces a simple example to illustrate a hard problem: shared meaning across markets. If something as basic as product descriptions differ by region, how can data products be reusable?

Botzen agrees it is difficult, pointing to the constant tension between central standards and local self-service. Even as a central team governs KPIs, people continue to create dashboards and new metrics on their own.

For Henkel, the path forward depends on establishing a central entry point that teams can rely on. Botzen frames the first question users should ask as “Is it already available?” and emphasizes that the organization should avoid duplicating efforts or building the same assets multiple times.

That is why the catalog matters, but it is not the only instrument. She also describes a SharePoint that covers business context, definitions, and guidance on how to work with data. The goal is to reduce duplication by making standards discoverable.

The data products that matter most: Finance, supply chain, sales

Some data products become foundational because so much of the business depends on them. Botzen says Henkel can see which data products are used most by plotting usage in the mesh, and the results are unsurprising.

“These are basically our SAP data products coming for certain domains,” she says, mentioning a finance data product, a supply chain data product, and a sales data product.

She describes them as “the core central data assets,” emphasizing how users build from these foundations by adding external data or domain-specific context.

Competitive advantage data: R&D, recipes, and customer feedback

When asked where Henkel holds a competitive advantage in data, Botzen points to R&D. She emphasizes the importance of R&D data products, highlighting the value embedded in recipes, formulas, raw materials, and the accumulated knowledge of what works and what does not.

This data is uniquely Henkel’s, which raises security questions. Botzen notes that it requires the company to think carefully about whether it is “secure enough to keep the data where it sits now,” especially during transformation and transfers between systems.

She also adds another competitive layer: Data from customer service and product feedback. In her view, the advantage comes from combining these streams to guide innovation and product evolution.

The biggest barrier beyond technology: Alignment and the people factor

Next, Schirmer asks about challenges beyond governance. Botzen’s answer moves straight to human systems. “The biggest challenge is the alignment and sticking to the alignment across those different groups of people,” she says.

But it also raises a broader question: are teams truly enabled to work with data in the right way? Botzen notes that for most employees, data ownership and quality are not their primary responsibilities. Their focus is on core objectives such as improving products, advancing R&D, and driving outcomes.

As a result, data often becomes a secondary priority for teams already deeply engaged in their day-to-day work. She describes this as a persistent challenge, one that Henkel is addressing through cultural initiatives, training programs, and active listening to understand where friction exists.

Botzen mentions asking teams why they cannot do the job as expected and what would help. “The people factor is definitely the bigger nut to crack than the technology,” she says.

Data culture as the “social glue”

While Henkel does not rely on formal incentive systems to drive ownership of critical data products, the organization works to formalize responsibilities through job descriptions and clearly defined roles for data owners, stewards, and system owners. But for Botzen, alignment on paper is not enough. What matters is whether people understand the value of data work beyond their immediate domain. As she puts it, it starts with the why. But building that understanding takes time, requiring continuous listening, explanation, and reinforcement to create lasting clarity.

That is why, alongside governance, Henkel prioritizes data quality and data culture. Botzen frames culture as the connective tissue that makes the operating model work. “The data culture is super important,” she says. “This is the social glue for bringing data to the people and getting something out of data.”

CDO magazine appreciates Dr. Katrin Botzen for sharing her insights with our global community.