Big Data in Italy: Market Up 24%, Yet Few Use Data to Decide

On 6 November 2026, Automazione Plus reported the findings of research by the Politecnico di Milano's Data & Decision Intelligence Observatory on how Italian companies use Big Data. The main finding is a contrast: spending is growing at a record pace, but the ability to turn data into decisions remains limited.

The research in numbers

The available information shows the following:

• Italian spending on Data Management & Analytics services and solutions reaches €5.07 billion in 2026;
• this is a record increase of 24% over 2025;
• only 18% of large companies have a Decision Intelligence approach, meaning they actually use data in decision-making processes: roughly 1 company in 6;
• around half of large companies have defined a Data Strategy.

The market is racing ahead: the ability to use data in decisions is not.

What this means in practice

The gap between having a data strategy and actually using data to decide is the most interesting point. If about half of large companies have formalized a Data Strategy, but fewer than one in five has a Decision Intelligence approach, then setting goals and investing in platforms is not enough. The intermediate steps are often missing: reliable data, processes that consume it, and people who turn it into action.

It should be noted that, for now, these are the only figures available to us: the market breakdown by component, the survey sample and any manufacturing-specific figures have not been verified. Anyone wishing to cite further details should consult the full article or the Observatory's report.

Why it matters to those running a plant

For production directors, plant managers and IT/OT managers, the message is cautious but useful. Growing analytics budgets do not automatically produce better decisions on the shop floor. In a factory, value emerges when data is reliable, contextualized (linked to machine, order, lot and shift) and tied to a specific decision.

OEE is one example: if downtime is recorded manually or without reason codes, the indicator says little and drives no action. If instead it is derived from signals acquired directly from machines and associated with production orders, it becomes a tool for deciding where to intervene. The same applies to downtime and traceability.

A practical approach to start from

In light of this scenario, a few operational recommendations:

• start from a few recurring decisions (for example, which stoppages to reduce first or how to reschedule an order) and identify the data they require;
• make sure shop-floor data is collected automatically, with adequate machine connectivity, and enriched with production context;
• use systems such as MES and data collection platforms to link plant signals to orders, lots and reason codes;
• measure whether the data really changes behavior: department meetings, priorities, maintenance interventions;
• extend advanced analytics only after this foundation is consolidated.

In summary

The €5.07 billion estimated for 2026 shows that interest in data is now structural. The 18% figure, however, is a reminder that competitive advantage does not depend on spending, but on the ability to deliver reliable information at the moment of decision. In manufacturing, this almost always starts with shop-floor data.