Why is supply chain AI still stuck in pilot mode in Europe?
Most buy the engine, few connect the pipes | Zygital8
Why is supply chain AI still stuck in pilot mode in Europe?
Most buy the engine, few connect the pipes | Zygital8
A survey of 100 European supply chain leaders at companies with $100M+ in revenue found that only 15% have fully rolled out AI. The report points to "data plumbing": getting suppliers, plants, carriers and customers to share data. So what is actually holding AI back?
Top barriers cited (multiple answers allowed)
67% poor or fragmented data · 61% integration with existing systems · 35% unclear use cases · 33% talent shortage · 30% ROI / scaling · 22% strategy gaps
If not big AI projects, where is the supply chain tech money currently going?
Into seeing clearly. The share of European companies spending over 0.5% of revenue on supply chain tech rose from 18% in 2024 to 26% in 2025, and the biggest jump went to tools that turn scattered data into one readable screen (called Business Intelligence): from 17% to 38%. Think of it as a control panel. Instead of piecing together spreadsheets from purchasing, warehouses and carriers, a manager sees in one place which delivery is late and which stock is running low. It is a smart order of operations: you cannot automate what you cannot see, so companies fix the view first and hand over the decisions later.
Why is supply chain data integration still broken after 20 years of ERP systems?
Because the problem isn't new, only the stakes are. 67% of European supply chain leaders name poor or fragmented data as a top AI obstacle, and 61% name outdated systems that don't connect (respondents could pick up to three). For years, companies have bought separate tools for purchasing, warehouses, transport and planning. The ERP, the central system that runs orders, stock and invoices, only holds part of the picture. The rest sits in other tools and in suppliers' portals, and none of them talk to each other. The report calls fixing this "data plumbing": the pipes that connect suppliers, plants, carriers and customers.
How are the top 15% structuring AI projects to force a measurable impact?
Only 15% of surveyed companies have fully rolled out supply chain AI. The report does not say what they did differently, but it is clear on what it takes to get there. It recommends starting from a clear operational problem, like improving demand forecasts or speeding up daily decisions, rather than from the technology itself. Early tests should be short and judged on what they teach, such as how ready the data is. Only the projects that prove their business value should be scaled, with a clear owner accountable for the financial results. Behind it all sits one idea: orchestration, the layer that connects suppliers, plants, carriers and customers so that the data flows between them rather than staying trapped in each tool.
The final take: so AI + messy data = a very fast mess...
Hold on a sec. If we're going to talk about plumbing, let's talk about plumbing in houses to get the idea. Nobody moves in before checking the pipes, right?
Same goes for AI: you check the data first. The AI is the thermostat on the water heater: it decides how hot things get, but if the pipes are clogged or leaking, it just pushes bad water faster. Same for companies: once the data finally flows where it should, someone in marketing will call it the fancy word "orchestration".