Building Data Culture from the Ground Up: Henkel’s Florian Roscheck on Grassroots Transformation
Thu 29 May 2025
Culture, not just code, is often the real catalyst for enterprise data transformation. At Big Data & AI World Frankfurt, taking place on 4-5 June at Messe Frankfurt, Florian Roscheck, Senior Data Scientist at Henkel, will bring a grassroots perspective to building data fluency from the ground up.
From internal learning communities to corporate tech blogging and corporate influencer programmes, Florian’s grassroots approach blends technical leadership with human-centred change. In this interview, he unpacks the tools, habits, and cultural nudges that are helping organisations turn curiosity into capability.
– – – – – –
What inspired you to speak at Tech Show Frankfurt this year?
I’m a huge fan of personal exchange and conferences for learning. I always take a lot away from them, and in the past years, I’ve brought many of those learnings back to the Henkel world—my workplace. Over the years, you start to realise that this whole data ecosystem only works if we all contribute. If you don’t give back, something’s missing.
So I want the full experience—the fun and the insight. That’s why I’m looking forward to sharing some of my thoughts, hopefully sparking some conversations, and gaining new insights at the conference.
Why is now the right time to talk about grassroots data culture?
The first question you have to ask is: What is a grassroots data culture? To me, it’s change that originates from the bottom, not from the top — from the doers, not just the managers or orchestrators. It’s about empowering those on the ground. I believe it’s never been easier to get started with AI or generative AI than it is today, because these tools are now accessible to so many people, not just those at the top.
Industry consultants like Boston Consulting Group recommend that 70% of successful AI transformation should focus on people and processes, and only 20% on the technology itself. So the question becomes: how do you fill that space? How would you do change management? How do you reimagine the processes, and how do you leverage the AI talent that you have in your organisation?
We often assume that top-down, centralised change is the most important. But when transformation is happening from the grassroots, which is the case now with AI, we need to meet that change where it is. The AI tools are now in the hands of everybody, and domain experts are becoming AI experts. This is why I think now is the time to talk about grassroots transformation.
What are some of the most effective grassroots tactics you’ve used to build data literacy and engagement across an organisation?
Certainly, I’m going to be talking about three initiatives I’ve either led or been part of. First, there’s the Henkel Data Club—our internal data community of practice. It’s mixed-level and currently has about 1,000 members.
Second, we run a data and analytics blog. It gives our internal experts a platform to share their knowledge, but it’s also a space for external data and analytics professionals, who might one day become our colleagues, to discover the technology we use and how we think about data.
And third, which is pretty new, we’ve launched a corporate influencer programme. We’re using it to attract data and analytics talent, and I’ve also seen how it sparks important conversations about the role of data and analytics in the company.
How can internal content help accelerate data maturity?
Yeah, that’s a great question. When you look at recent surveys like the BARC Data Culture Survey, you’ll see that around 40% of companies cite a lack of employee knowledge about data and analytics as a major obstacle, and about a third mention a lack of communication.
So, what do you do if people lack knowledge? You give them knowledge. If there’s no communication, you communicate. It’s pretty straightforward—and that’s where internal content comes in.
You need both content and a distribution platform. In my experience, what really works is sharing real-life stories from your business—people solving real problems using your own data tools. It’s easy to read best practices online, but applying them to transform your company is the hard part.
In the Henkel Data Club, I’ve seen people across the business get inspired by shared content. They generate their own business benefits and bring those stories back into the community. It legitimises experimentation, and I think that’s how change spreads.
From your experience, what are the key cultural or organisational blockers that slow down data product development, and how can teams overcome them?
That’s such a great question. I think when you work in an enterprise, you’re very much tuned for efficiency and for getting the best out of existing processes. But there is a challenge because say this world of data, especially with the rise of generative AI, becomes far more open and wide when you think of the possibilities of GenAI.
Suddenly, innovation is falling into the hands of technologists—people who were often seen as service providers a decade ago. Now, they’re confronted with changes that can truly impact the business.
One major blocker I see is staying stuck in an execution mindset—focusing only on solutions instead of spending time in the problem space. It’s crucial to understand what the business problem actually is and how we can help solve it. So what you need there is you need to build bridges, and this takes resources, time, and a concerted effort.
What advice would you give to companies trying to create a data culture from the ground up, not just top-down?
You need to dedicate resources. This isn’t something you can do strictly organically. For example, setting up the Henkel Data Club and operating the data and analytics block that we have takes time. That means setting aside a budget, not for direct project work, but for upskilling and capability-building.
This change is fundamentally important because this change is now in the hands of domain owners across the business. If you build a data culture, you have a means and vehicle to scale this change across your organisation. This is why it’s so important.
Another thing that I’ve found for us: we have given our data and analytics experts free space, let’s say, to be geeky and to live their passion. This fits in well with the data culture ecosystem because when you have a platform like the Henkel Data Club in which these experts share knowledge, you not only scale war or success stories from the business, but you can also scale best practices across the organisation. This then enables progress at large in niches in the organisation where you wouldn’t get to otherwise.
What’s your view on the strengths and weaknesses of the DACH region when it comes to AI experimentation and scale-up?
Yeah, this is a great question. Just as a bit of background—I worked seven and a half years in the US, in a mid-sized company and a startup, and I’ve also worked in two large German corporate environments. When people talk about weaknesses, they often mention regulation, compliance, and financing.
But actually, I want to focus on culture and education. I’ve made some observations, especially in the German working culture, and particularly in corporate environments. By the way, if you’re listening and this resonates with you, I’d love to connect—let’s have a conversation.
There’s a fear—fear of backlash, of trying something new, of hearing negative feedback, of breaking the rules. In my view, in Germany, people often wait for someone to tell them the rules. And while Germany has proven itself to be great at innovation, we’re now in a space where the rules haven’t been fully written. Sometimes, no one can tell you exactly what they are.
So we fall back on thinking: “I fulfilled the requirements, followed the process, didn’t stand out—so everything must be fine.” But that’s at odds with what innovation really needs. You need to do something no one’s done before. And when no one can tell you how to do it, you need courage.
I don’t think we value courage as much as other cultures do—certainly not as much as in the US. There, it’s more common to think: “Do what you want, but own the outcome.” If you don’t know something, try it. If a process doesn’t fit what you’re doing, you adapt it.
In the US, it feels to me as if people are more accustomed with winning and losing all the time—there are more ups and downs, and that creates more recognition for being bold. Especially in innovation, and particularly in data and analytics, I think courage is really the key. In cultures that value conformity, that kind of courage can be harder to nurture—but it’s essential.
What’s one underappreciated technique or trend in the AI/ML space that more teams should be paying attention to?
I think there are two ways to look at this—tools and technology on one side, and people and processes on the other.
On the technology side, think about the requirements people now expect: tools that value privacy, are cost-efficient, and don’t require huge investments. The scope for AI use, particularly with agent systems, is becoming more focused—smaller tasks, more specific outcomes. So in this context, small language models (rather than large ones) can be particularly effective. They’re cheaper to run, can be deployed privately on your own infrastructure, and are often enough to do the job.
On the cultural side, there’s a trend from Silicon Valley and the product world that’s catching on in tech: empowered teams. This is especially trickling down in the tech sector. An empowered team is trusted to solve a business problem, not just follow orders.
Say someone asks for a red button on a dashboard. If you’re just building that button, your job ends there. But if you’re empowered—if you know how technology like AI works, and what’s possible—you might help discover a better solution together with the business partner. That’s the shift: from execution to discovery.
When you, as a technologist, are working together with a business person in the discovery process to find out what that problem is, you feel fully empowered. This is, in a nutshell, what empowered teams is about, and I think getting more of this culture into our companies helps us accelerate this change and also be more courageous.
You’re going to be speaking at Tech Show Frankfurt. Is there anything you want to tell our audience?
My presentation is about data culture with the grassroots approach—I’ve thought a lot about what I can give people. We’ve invested so much effort into building the Henkel Data Club, publishing our blog, and launching the corporate influencer programme.
So I’m going to share very tangible lessons from that journey—things we learned the hard way—so others don’t have to go through the same challenges. I want people to walk away with actionable insights and the confidence to get started right away.
Don’t miss Florian Roscheck’s session at Big Data & AI World Frankfurt as part of Tech Show Frankfurt, where he’ll share practical lessons from inside Henkel on building grassroots data culture, turning curiosity into capability, one community at a time.
Note from the Editor: All quotes are verbatim from a recorded interview, lightly edited for clarity.
