Almost every manufacturer is working with AI, yet most have little to show for it on the bottom line. That's rarely down to the technology. It's down to how it's rolled out: a tool bolted on while the work itself stays the same. Here's what actually works.
The figures are strikingly consistent. Nearly eight in ten companies now use generative AI, yet roughly the same number see no measurable effect on their profits. In manufacturing, the vast majority are still in the exploration or pilot stage, and only a handful have truly woven AI through their whole operation. Plenty of motion, then, but little gain. The question is: how can that be?
In our conversations with manufacturers we keep hearing the same thing. There's a trial with a smart chatbot, someone has bought an AI licence, quotes are written faster now. All true. And yet, at the end of the quarter, there isn't a cent more in the bank and not an hour of work less. It feels modern, but it delivers nothing.
That's not bad luck, and it isn't the wrong tool. McKinsey's research among operations directors in manufacturing shows that around two thirds of companies are still in the exploration or targeted-pilot stage, and that only about 2 percent have fully embedded AI in their operation. The ambition is there. So are the budgets. What's missing is the step from "we use AI" to "our work is done differently".
This, to our mind, is the heart of it. Most AI projects lay a smart layer on top of a process that otherwise stays untouched. The estimator uses AI to type a text faster, but still hands the quote to work preparation by hand. The data is still retyped from one system into another. The AI makes one small step quicker, while the chain around it creaks just as hard as before.
McKinsey sees exactly that: the companies that do make money from AI are the ones that redesigned the way they work while introducing the technology. Not a tool added on, but the process rethought. That takes effort. And that effort is done by people, not models.
AI only truly lands once it's in your process, not just beside your people. That's where calm and gain are born.
Our approach is therefore built around two things at once: the people and the process. We sit down beside your staff on the work floor, first watch how they really work, and set up sharp improvements right there. We help where the work actually happens, so that it sticks.
At the same time we look at the chain around it. Which step costs needless time? Where is information entered twice? Where does work pile up because no one has the overview? That's where we build a solution that runs on top of your existing systems. You don't have to replace your ERP or your bookkeeping. We make what you already have work together better.
The difference is in the order. People first, then the process, and only then the technology. Not the other way round. Technology you unleash on people without understanding their work gets ignored or clicked away. Technology that makes your people's work lighter stays.
An example from our own practice: for an organisation with a fragmented grant administration, scattered across countless separate places and partly still on paper, we first mapped the ideal process and then built a single platform where everyone involved works together. Applications are now fully digital, there is one version of the truth, and nothing falls through the cracks. That saves hours of work every month and, just as importantly, a great deal of noise and hassle. You can read the whole story in our case study.
The same principle works on the floor at a smaller scale. An employee who retypes data every day, keeps an eye on orders and maintains reports effectively gains a virtual colleague that takes over the repetitive work. The person stays in charge and in control. The time that frees up goes to the work that really matters.
You don't need to set up a six-month programme for this. In fact, that's precisely the trap. Begin with the one process where the pain is greatest, put a first improvement in place beside your people, and measure what it delivers. If it works, you scale it up. If it doesn't, you've lost little. That's how you build trust, with your people and in the numbers alike.
The manufacturers that will be making money from AI a year from now aren't the ones with the most expensive licences. They're the ones who dared to redesign their work. That doesn't start with a tool. It starts with a conversation about where your people lose time every day.
Why does AI in manufacturing often not deliver a profit yet?
Because the technology is usually laid on top of an unchanged process. One task gets faster, but the chain around it stays the same. Gains only appear once you redesign the work itself, and that's done by people, not the tool.
Do we have to replace our ERP or production systems to get started with AI?
No. We build a layer on top of what you already use and make your existing systems work together better. You don't have to replace anything to save time.
How quickly do we see results?
Fast. We start on the work floor beside your people and often set up the first improvements within days. A short scan shows where the biggest gains are, without having to launch a large programme.
Does AI replace our people?
No. We put AI right beside your people, so they're freed from the repetitive work and gain time for the work that really matters. The person stays in charge and in control.
We'd be glad to come and take a look on your work floor and give you an honest, concrete picture of where the biggest gains are and how we can help.