Beiersdorf is using a structured R&D data platform to help reformulate iconic skincare products for a lower environmental footprint, while preserving the safety, performance, and sensory experience consumers expect.
For Beiersdorf, sustainability is not an initiative running alongside product development. It is becoming a core formulation constraint.
The maker of Nivea, Eucerin, Aquaphor, Coppertone, and La Prairie has committed to achieving net-zero emissions by 2045. That target puts new pressure on R&D teams to rethink ingredients, processes, packaging, and product design without compromising the product performance and skin feel that have made its brands household names.
In a recent C&EN webinar sponsored by Uncountable, Beiersdorf R&D leaders Thomas Schornstein and Yanel de Morel discussed how the company is using structured experimental data, machine learning, and global collaboration to support more sustainable product development.
How the maker of Nivea and Eucerin is using structured R&D data to pursue net zero without changing what its products feel like on skin
.png)
Beiersdorf faces a challenge that many R&D organizations would envy, and few would want to solve: its products are so well known and well loved that changing them carries real risk. Yet change is unavoidable.
The company behind Nivea, Eucerin, Aquaphor, Coppertone, and La Prairie has committed to reaching net-zero emissions by 2045. For its R&D teams, that means reformulating products, reassessing ingredients and packaging, and reducing environmental impact, all while preserving the safety, quality, performance, and sensory experience consumers expect.
That tension was central to a recent C&EN webinar, moderated by contributing editor Kelly McSweeney and sponsored by Uncountable. The discussion featured Thomas Schornstein and Yanel de Morel from Beiersdorf’s R&D organization, alongside Uncountable co-founder and chief revenue officer Will Tashman.
“One of the main challenges that we are facing is transforming our iconic products, as they are well known and also loved by our consumers,” Schornstein said. “So how can we give the same product experience with a better environmental footprint?”
Sustainability becomes a formulation constraint
For de Morel, who joined Beiersdorf’s Americas Innovation Hub in Mexico in 2017 and progressed from prototyper to team lead, sustainability is not a separate corporate agenda. It is now a core R&D challenge.
Beiersdorf’s sustainability commitments span consumers, society, and the environment. The company received a triple-A rating from CDP this year, placing it among a small group of top performers from roughly 50,000 companies assessed. But the next stage of its commitments will be harder to achieve.
“We set the target of achieving net zero emissions by 2045,” de Morel said. “And as you can imagine, this will bring significant complexity to the work that we do, not only in R&D, but all of the processes in the company that we have. And therefore, we have to really think about what type of tools we can bring into the organization that will support this change.”
That complexity sits on top of constraints that were already difficult to balance. R&D must continue to deliver safe products, meet quality expectations, and satisfy consumer demand for innovation. Sustainability adds further variables across the entire product lifecycle, from the sourcing and processing of raw materials to how ingredients interact in a formula and how the formula performs alongside its packaging.
“We’re looking from sourcing raw materials, what kind of processes does raw material also go through?” de Morel said. “When we incorporate them in our formulas, how do they interact also within the formula? And then, how do they behave with the packaging?”
From the software provider’s perspective, Tashman summarized the shift simply. Developing a good product is no longer enough. Teams must develop a product that performs, can be made efficiently, and has a lower environmental impact.
A digitalization journey decades in the making
Beiersdorf’s effort to modernize R&D data management did not begin with AI. According to Schornstein, the company started its digitalization journey in the 1980s, building a centralized IT documentation system to track experiments and standardize raw-material information.
“However, when it comes to reporting, Excel was the tool to be used, and also to visualize the analysis,” Schornstein said.
For a company with more than 140 years of skincare innovation and R&D labs in Hamburg, Brazil, India, China, and New Jersey, the central challenge was not a lack of data. It was making that knowledge accessible and useful across a global R&D network.
“It’s particularly crucial for us that we share this wealth of knowledge and data as a global player within our R&D labs, not just in Hamburg,” Schornstein said.
Tashman argued that this is a familiar pattern across many R&D-intensive industries. Traditional laboratory information management systems can be effective at tracking samples and test results, he said, but they generally do not capture the full formulation or recipe context behind those results. Electronic lab notebooks preserve rich experimental context, but often function more like “scientific word processors” than systems for generating the structured, reusable data required for modeling.
The technical challenge is only one part of the problem. The other is adoption.
“People don’t like systems,” Tashman said. “They like their spreadsheets because it’s their system.”
From pilot project to global platform
Beiersdorf became one of Uncountable’s first customers in 2018, at a time when the software company still operated more like a data science consultancy. Tashman described the early working model as a “they send a spreadsheet, we send them back spreadsheet type of relationship.”
Around 2020, the relationship expanded toward a more integrated platform. For Beiersdorf, the value was not confined to improving a single formulation project. It was the ability to capture and reuse learning across projects, teams, and locations.
“One of the benefits that we’ve seen also during the pilot phase was that we could also leverage much more across projects, so not just on project level,” Schornstein said. “New tools like the Uncountable platform allow us also to derive general learnings out of our data of each experiment conducted worldwide, and to further connect the global R&D network.”
When asked how Uncountable compared with alternatives available at the time, Schornstein highlighted a particular challenge in chemical and materials science: useful models must often work with relatively small datasets.
“We’ve seen not many competitors being able to work also with AI machine learning within a low data environment, like it is common when you’re working in a chemical or material science area,” he said.
The practical result, Schornstein said, is that formulators can analyze and act on data directly rather than waiting for specialist data-science support.
Deriving general lessons from experimental data “doesn’t have to be conducted by a data scientist,” he said. “It can be also done by the formulator directly.”
That matters because formulation innovation rarely happens in isolation. “We do see sustainability and innovation as actually a team sport,” Schornstein said. “We truly believe that innovation also happens at the interface of disciplines and technologies.”
De Morel described a similar benefit at the bench level. The platform, she said, improves visibility into who is working on related projects and makes it easier for colleagues to connect, exchange knowledge, and collaborate.
AI as augmentation, not replacement
Both Beiersdorf speakers were measured in how they described the role of machine learning. The goal is not to automate formulators out of the process, but to give them a more capable way to navigate complex tradeoffs.
“The way we see it, using the AI machine learning model there is rather on augmenting the formulator’s capacity,” Schornstein said. “It’s always not just the model or not just the human, it’s rather a combination of both.”
Tashman framed the point in practical terms. A formulator may need to balance a dozen competing properties, from product performance and sensory characteristics to cost, ingredient availability, processability, packaging compatibility, and environmental impact. That is too much for anyone to optimize intuitively.
The challenge becomes even greater as sustainability targets drive the adoption of new materials with limited historical data. In that context, the platform is not a substitute for formulation expertise. It is a way to help experts explore a wider range of viable options and make better-informed experimental decisions.
De Morel shared the webinar’s clearest outcome metric from a pilot project she personally tested. The approach significantly reduced the resources required and enabled the team to identify a solution in the second or third development loop.
“Instead of doing a lot of batches, we have to produce 30% of the batches that we normally would do to find a solution,” she said.
She also emphasized a less measurable, but potentially more valuable, outcome: the ability to uncover promising experimental directions that a formulator might not otherwise have considered.
“We have had really nice examples where we are doing experiments that we don’t expect to do and that we would not have thought about,” de Morel said. “We were getting some recommendations from the platform and also exploring in these recommendations and having some nice prototypes in the way.”
Of course, not every unusual recommendation succeeds. Tashman acknowledged that when the system proposes experiments outside a team’s normal playbook, some will fail. But that is not necessarily a drawback. It is part of the purpose of R&D: testing ideas that are new, different, and not guaranteed to work.
What comes next
The conversation pointed to several areas where Beiersdorf and Uncountable see further opportunity: image analysis, consumer sentiment data, stronger connections between R&D and production-scale processes, LLM-based semantic search across structured experimental results, and continued investment in laboratory automation.
Still, Schornstein’s closing point was less about the technology roadmap than about the relationship behind it. Through the proof of concept, he said, Uncountable demonstrated that it could be a trusted partner. That trust made the subsequent, deeper IT integration possible.
As Tashman put it when the discussion turned to connecting systems and making enterprise-scale digitalization work: “It takes two to tango.”
See how Uncountable helps formulation teams centralize experimental knowledge, evaluate competing product requirements, and identify promising next experiments with greater confidence.
.png)
.png)
.png)
