At Henkel, applying a single global process can improve performance at one plant while constraining it at another. That tension shapes the company’s digital operations strategy.
In this interview, Benjamin Johnson, Global Head of Digital Operations, Adhesive Technologies at Henkel, explains how the company segments plants by digital maturity and decides where to enforce global standards versus allowing local flexibility. He also reflects on an early lesson around data quality, and how real-time visibility is designed to support, rather than replace, plant-level judgement.
What operational challenges emerged as Henkel scaled across a large number of plants?
The largest challenge in digitalising our operations has been, and continues to be, the sheer diversity and complexity of our manufacturing footprint.
We operate more than 120 adhesives production sites globally, spanning a wide range of equipment age, automation maturity, and infrastructure sophistication. Across these sites, we run over 100 major manufacturing technologies, each with distinct processes, and serve more than 800 industries with highly varied product and quality requirements.
Given this reality, a one-size-fits-all digital approach was neither practical nor commercially sensible. Instead, we developed a segmented digital strategy built around clear archetypes. This allows us to manage complexity, focus investment where it delivers the most value, and apply digital capabilities in a way that is appropriate for each site.
Our approach is structured around four archetypes that define the level of core infrastructure, data availability, and digital capability at each site. These are supported by a centrally governed catalogue of proven use cases that can be deployed and adapted locally as needed. This model provides the benefits of scale and standardisation, while enabling site-specific customisation to improve operational performance and support our broader digital transformation.
Where did efforts to standardise across plants require more adaptation than expected?
Standardisation is never easy and requires significant effort on the shop floor. This includes developing a practical understanding of processes, building buy-in from key stakeholders, and committing to sustained, long-term change management across our sites.
In practice, the challenge lies in the fact that a process that improves performance at one site can inadvertently constrain or reduce performance at another.
As a result, a key focus of our digital strategy is deciding upfront whether value will be created through a global standard, or by allowing sites to adapt within a governed framework.
For example, ensuring consistent product specifications and quality across our network requires rigorous global standardisation. By contrast, a real-time shop floor performance dashboard may deliver greater value when developed locally, so it can directly support site-specific priorities and address performance gaps where they matter most.
Which digital initiatives scaled easily, and which were harder to extend beyond pilots?
For us, the strongest indicator that a solution will scale successfully is not technology-related, but people-related. When we see strong pull and genuine buy-in from sites, we are more likely to secure the focus, ownership, and resources needed to build, iterate, and sustain a capability over time, rather than treat it as a short-term pilot.
The speed and effort required to scale is largely shaped by how much standardisation we enforce. Platforms that allow sites to use digital and data within a governed framework tend to scale more quickly. By contrast, solutions that depend on tight, end-to-end standardisation typically scale more slowly, although they can deliver stronger consistency where it is required.
In general, use cases at the Level 4 enterprise layer scale more easily because of higher levels of standardisation, simpler data models, and fewer system and stakeholder dependencies. As we move further down the automation stack into MES, control systems, and sensor layers, system complexity increases. Process variation and data diversity also grow significantly.
For full-stack architecture use cases, we assess carefully where standardisation creates the most value and where flexibility delivers better outcomes. In some cases, allowing freedom at the OT layer to accommodate local integrator ecosystems and site realities is more effective than mandating uniform infrastructure, vendors, or architectures that may slow adoption or reduce local effectiveness.
What early assumptions about data and system inputs had to be revised?
Earlier in our journey, we underestimated the level of effort and sustained focus required for effective input and data management. We assumed that if we built the right systems, input quality would improve over time through regular use.
We learned that this assumption was flawed. Input quality is foundational. Without getting it right, it is difficult to make meaningful progress in change management or to consistently realise business value from digital initiatives.
As a result, for all new digital capabilities we now invest time upfront in designing for high-quality inputs. This includes using automation where possible, building error-proofing mechanisms, and enabling users to quickly identify and flag data quality issues when they occur.
Since adopting this approach, we have seen stronger user adoption, faster time to value, and higher overall user satisfaction. This reinforces the importance of treating data quality as a core design principle rather than an afterthought.
How has real-time visibility reshaped central and plant-level decision-making?
Improved data use and greater operational visibility have changed the way we work, with a large proportion of our core processes now using digital tools to support day-to-day and strategic decision-making.
We have defined a clear infrastructure and analytics strategy aligned to the processes it supports. For example, our operational platforms provide real-time shop floor dashboards that enable operators to identify and respond to issues quickly, helping ensure daily performance targets are met. In parallel, our central data platform is designed to support management and leadership, enabling decisions to be steered, progress against longer-term objectives to be tracked, and resource allocation to be prioritised across weekly, monthly, and annual operating rhythms.
This approach has improved transparency, accountability, and operational responsiveness. It also ensures that digital and analytics investments are directly connected to business outcomes and aligned to performance objectives.
Which manufacturing decisions still benefit most from human judgement rather than data?
People remain at the centre of our digitalisation journey. More than 10,000 employees are directly involved, and sustained progress depends on building new skills, developing digital fluency, and fostering a culture of continuous learning and improvement.
Given the diversity of our operations, our ambition is not to replace human judgement, but to support it. By equipping teams with timely and unbiased data, we aim to strengthen decision-making at every level and enable people to apply their expertise more effectively.
At its core, our strategy is straightforward: use digital execution tools to reduce time spent on manual and administrative work, and reinvest that time, supported by better data, into problem-solving, continuous improvement, and operational performance.
