Twelve months after my appearance on the LogTech Podcast, the world of logistics software has completely changed. But it seems very few people care.
Why should it? The numbers are good; everything’s going as usual.
1,500 person-days for a WMS implementation. Fifteen years ago, it was 650. Same result.
I mentioned this figure just under a year ago on the LogTech Podcast, in the episode “Straight Talk with André Kaeber: On Mindset, Innovation, and Dead Horses in SaaS.”
I listened to it again before writing this text. And sorry: not a single sentence in it is outdated.
That’s exactly the problem.
Twelve months in which everything has changed. Except for us.
Since this recording was made, the way software is developed has completely changed.
Building software has become inexpensive. Coding agents write production code; humans review, make decisions, and take responsibility. What used to be a year-long project is now a matter of weeks. We build this way ourselves, every day. Yard management, transport management, interface connectors, agents, bots…
Systems communicate with each other. Open protocols like MCP and A2A turn the interface into what it should always have been: a formality. The interface used to be the most costly argument against interchangeability. That argument is now crumbling.
Agentic AI has moved beyond the demo phase and is now in full operation. Physical AI makes old systems readable without anyone having to retrofit them. And AI is no longer just an add-on feature. It’s infrastructure—just like electricity, just like railways, just like the ground on which the factory building stands.
Twelve months in which almost every technical excuse has lost its basis.
And yet we keep talking about the same five things. Because they were never technical.
1. Sales isn’t your problem. Your operating system is.
When growth stalls, the reaction is always the same. Pressure on sales. Pressure on marketing. A new campaign, a new tool, a new quota.
That’s a lie that feels good. Because failure rarely stems from a single department. It stems from the target operating model—the company’s internal operating system.
In SaaS, this becomes painfully obvious. Pressure from venture capitalists forces “hockey stick” growth. So the founder brings in customers who aren’t a good fit for the product. Next comes customization, which dismantles the scalable model. The end result is a company operating in silos, paying the price for its own complexity.
“It’s never just one issue or problem in a department. Actually, the entire operating system needs to be reviewed.”
If you have a dead operating system, you can generate leads until you drop. You’ll still be overwhelmed.
2. The Productivity Paradox: You’re paying for stagnation in a glossy package.
Twenty years of system implementations. WMS, TMS, yard management—all of them implemented two or three times over. According to Gartner, productivity is stagnating or declining. We’re running on a very expensive treadmill.
This is most clearly evident in the amount of work involved. 650 person-days 15 years ago. 1,500 today. For the same result.
Then there are the modernization releases: millions spent just to maintain the technical status quo. No new processes, no new customers, no new euros. Just a system version that’s supported again.
A year later, this calculation has become even more embarrassing. The cost of building software has fallen. The quotes for implementation have not. This gap isn’t a market price. This gap is a bet that you won’t do the math.
Check your calculations.
3. Comprehensive Insurance Mindset: The Fear of Demolition.
Innovation requires an open mindset. Instead, our economic culture is obsessed with security. In Germany, we insure against things we won’t even live to see ourselves—our own death, for example.
We apply this “full coverage” approach to the IT landscape. We treat legacy systems as if they were sacred. They’re a millstone around our neck.
Sido built an entire song around this same radical idea, “Pyramiden”: The old must go so that the new can take its place.
“In my opinion, you just have to cut out a whole lot of it. Tear it out and start over.”
The usual objection to this has always been the price. Demolition is expensive, new construction takes time, and the risk is too great. That very objection has become obsolete in the space of twelve months. New construction today costs a fraction of what it did in 2024. Anyone who’s still painting the old house isn’t doing it out of reason. They’re doing it out of habit.
Transformation means laying bare the foundation. And, if in doubt, bringing out the wrecking ball.
4. Status symbols trump innovation. And middle management is holding things back.
Resistance rarely lies at the base. It lies at the level above it.
It’s not about processes there. It’s about the company car, the salary, and power through meetings. Control via meetings isn’t leadership. It’s a calendar that pretends to be.
Automation does not threaten the scheduler. It threatens the relevance of a management level that does nothing but delegate and administer. An agent needs neither status symbols nor coordination meetings.
And it makes the dispatcher stronger, not redundant. His role is shifting. Less typing, more decision-making. Fewer fields to maintain, more exceptions to resolve, more rules to set, more responsibility to bear. That’s an upgrade, not a cost-cutting measure. But it only happens if someone defines, fills, and trains for this new role beforehand. That’s exactly where most companies fall short.
Stop treating employees like data-entry drones. And as a leader, ask yourself honestly whether you’re still adding value or just tying up resources in endless meetings. Anyone who blocks change to protect their own interests is signing a death warrant in installments—for everyone in the company.
5. Build to Evolve: From Monoliths to Orchestration.
The era of rigid, monolithic ERP systems is over. The future belongs to orchestration. Software works like a modular puzzle. We call this “Build to Evolve.”
The new KPI isn’t a silver bullet. The new KPI is redundancy and interchangeability. In a volatile world, what matters is the ability to switch logistics providers or system modules on the fly. That’s not an architectural luxury. It’s a matter of survival.
The tools for this are already in place. Agentic AI makes autonomous decisions across system boundaries. Physical AI makes existing equipment “readable”: The U.S. startup Guidewheel measures the condition of ancient machines not through expensive sensors, but through power consumption. This makes every old object accessible. And affordable energy is the real driving force behind it all, because otherwise computing power becomes a locational disadvantage.
What was just an idea a year ago is now a work in progress. Open protocols ensure that modules can communicate with one another without turning into an integration project every time. Interoperability used to be just a slide in a presentation. Now it’s a decision.
George Bernard Shaw put it succinctly over a hundred years ago: The reasonable person adapts to the world; the unreasonable person insists that the world adapt to him. All progress depends on the unreasonable person.
We’ve been very sensible when it comes to logistics for the past twenty years.
Conclusion: The Courage to Make a Radical Cut
Logistics innovation isn’t just a software feature you can add on. It’s a matter of mindset.
If you digitize inefficient processes, you’ll end up with inefficient digital processes. Faster but wrong is still wrong.
A dead horse won’t come back to life just because you saddle it with an agent. It’ll just cost more.
Away from managing shortcomings. Toward creativity. That means a spirit of experimentation, radical simplification, and the realization that redundancy is a strength today.
Technology has been evolving for twelve months. We haven’t. We’ve run out of excuses, and the dead horses are still in the stable.
Are you ready to get out before the market forces you out?
Nothing ventured, nothing gained. Almost always!
If you don’t take a chance, you’ve already lost!
