People, data, technology — the only change management order that works in manufacturing
Most digital transformations in UK manufacturing start with technology, struggle with data, and only address people when adoption is already failing. The progressive ones do it in the exact opposite order. Here is why — and what the change management plan should actually look like.
The technology lands. The people don't follow. The benefit doesn't show up.
The pattern is so consistent that it has become predictable. The MES is deployed on time. The PLM goes live. The IIoT layer is connected. Three months later, the operations director admits that the operators are still using the spreadsheet, the engineers are still emailing PDFs, and the dashboards on the wall in the canteen are showing data that no one trusts. The technology is fine. The change has not happened.
Change management is not a slide. It is the work that converts deployed software into used software.
"You don't have a technology problem. You have a people problem that the technology was supposed to solve, and didn't, because no one did the 'people' work."
Why change management matters more in manufacturing than anywhere else
Manufacturing is a different change-management discipline from the back-office and customer-facing transformations that dominate the consulting literature. The user is on the shop floor, often mid-shift, often holding a torque wrench, often three steps from a piece of equipment that does not pause while they learn a new system. A workflow that takes fifteen seconds longer than the previous workflow does not survive contact with a real shift. Prosci's change management research — the largest body of comparative evidence in this field — finds that the projects with "excellent" change management are six times more likely to meet objectives than projects with poor change management. That ratio is, if anything, conservative on the shop floor.
And the timing matters. MIT Sloan's ongoing research on digital transformation consistently finds that organisations with mature change-management capability outperform peers on both adoption and time-to-value. UK manufacturing, with its tight margins and skilled labour shortages, cannot afford to be on the wrong side of that curve in 2026.
People, data, technology — in that exact order, and here is why
People first — because they decide whether the rest of the programme exists.
The first conversation is not about software. It is with the operators, team leaders, engineers and supervisors who will live with the result. Why is the current way of working frustrating? What would "better" look like for them? Where do they not trust the existing data? Where do they routinely work around the official system because it does not match the reality of their job?
These conversations do two things. First, they generate the design input that turns a generic deployment into a fit-for-purpose tool. Second, and more importantly, they create the early adopters who will champion the system through the awkward first weeks. The shop-floor team leader who was asked her opinion in week one is the team leader who pushes back on the cynics in week six. That is not a soft benefit. That is the single most important predictor of adoption.
The people work happens on the shop floor, not in a workshop room with a flip chart.
Data second — because trustworthy data is the bridge between people and technology.
Once the people are engaged, the next question is: do they trust the data? If the operator looks at the OEE dashboard and immediately says, "that number is wrong, we did a changeover at 10:30 and that's not reflected," the whole programme has lost credibility before it has begun. Data trust is not the same as data perfection. It is the result of a deliberate process: defining what each metric means, naming the owner who is accountable for it, building the feedback loop that lets operators flag anomalies, and acting visibly on those flags.
Data trust is built one query at a time. The first time an operator asks, "Why does the dashboard say we made 40 units when we made 43?" and gets a sensible answer within a day, the system has just earned a unit of credibility. The first time the same question is answered with a shrug, the system has lost ten units.
Technology third — because by now, you actually know what to build.
Only after the people work and the data work has shaped the requirements does the technology choice make sense. The PLM module, the MES workflow, the IIoT dashboard — they are now configured to the workflow the team actually uses, populated with data the team actually trusts, and championed by the team leaders who shaped the design. Adoption is no longer a hopeful afterthought. It is the natural consequence of the work that came before.
This is the order the Cranfield School of Management research on operational change has consistently identified as the high-success pattern. It is also, in my consulting experience, the most reliable way to convert digital investment into business value.
What "people first" actually looks like on the shop floor
The phrase "people first" gets used so loosely it has become a corporate platitude. In manufacturing, it has a concrete operational meaning.
It means the engineering manager and the maintenance team lead are in the requirements workshop before the software vendor is. It means shop-floor walks happen before the data model is drawn. It means the "day in the life" of an operator on a Friday night shift is documented before any screen mock-up is made. It means the change-impact assessment names individual roles, not just functions, and says specifically what changes for each one.
It also means a visible structure for raising concerns and seeing them acted on — not a feedback form, but a weekly stand-up where shop-floor input is shown to land in the next sprint. The credibility of the whole programme is built or lost in those stand-ups.
What we would tell you if we were sitting across the table
Start change management before the technology selection, not after the deployment.The people input shapes the requirements. Try to bolt it on later and it becomes damage control.
Name your shop-floor champions explicitly. Two per shift per area, paid time built into their week, listed on the project plan. Not volunteers in their spare time.
Build data trust before you ask people to act on the data. A "why does the dashboard say…?" question answered fast wins more adoption than any training session.
Train in context, not in classrooms. Twenty minutes on the line beats four hours in a meeting room every time.
Make leadership visibility part of the rollout, not optional. The operations director walking the floor in week one, asking questions and acting on what is raised, is worth a year of internal comms.
Measure adoption alongside throughput. What gets measured gets managed. If adoption is not on the weekly steering pack, it is not really being managed.
Recognise the early adopters in front of their peers. Public recognition for the operators who got there first is one of the cheapest and most effective change-management levers in manufacturing.
The bottom line
Digital transformation in manufacturing is fundamentally a people transformation that happens to need data and technology to land. The order matters: people first, data second, technology third. Get the sequence right, and the technology takes care of itself. Get it wrong and the best software in the world will not save the programme.
If you are about to start a digital programme and the change-management plan is a single line in the project schedule, that is the line to fix first.
Coming soon — The TJ Manufacturing Interview Series. Tim Shelley in conversation with Gavin Hill, Head of Information Management Research at the AMRC, on the cultural side of digital adoption and the practical things UK manufacturers can do to bring their people along on the journey. Get notified when it launches →
Planning a digital rollout and worried about adoption?
Our team has led adoption-first manufacturing rollouts across UK SME, mid-cap and multi-site operations. Half a day, your operations leadership, a whiteboard, and a credible change-management plan at the end of it.