Digitalisation does not mean turning a paper form into a PDF or buying a new platform. It means redesigning a process so information, decisions and responsibilities move in a clearer, measurable way. The goal is not to introduce technology everywhere, but to choose where it can deliver a verifiable improvement.
A journey still under way for many SMEs
According to Eurostat, 71% of EU SMEs had reached at least a basic level of digital intensity in 2025, while the European target for 2030 is above 90%. The 2026 Digital Decade report recognises Italy’s progress in business digitalisation but still identifies gaps in digital and ICT skills.
These figures describe a direction, not the priority for an individual business. To decide where to begin, examine day-to-day work, the systems already in place and the information currently lost, duplicated or reconstructed manually.
Identify the bottleneck
Who receives the request? Where is data copied? Which approvals block the workflow? How many exceptions are handled outside the system? Re-entered orders, quotes scattered across emails and spreadsheets, unassigned tickets and manually assembled reports are useful warning signs.
Before intervening, distinguish the cause from the symptom: automating an unclear process can make errors happen faster too. The map should include people, systems, input data, decisions, exceptions and the expected outcome.
Choose a measurable priority
A strong initial candidate combines frequency, manual effort, the impact of errors and the ability to measure the outcome. Customer enquiry management, order intake, document approvals, report generation and synchronisation between CRM and management systems are common examples, but priorities depend on the business.
Before the project begins, establish a baseline: average time, number of steps, corrections, delays and manual activities. Without starting data, it is impossible to distinguish a positive impression from genuine improvement.
Integrate what already exists before adding tools
Businesses often already have a CRM, ERP, e-commerce platform or specialist software. The problem is that information remains siloed. A connector or dedicated service can transfer the necessary data while applying validation and business rules, without immediately requiring every platform to be replaced.
Artificial intelligence should also be introduced only where it adds value. It can help classify requests, extract information or prepare content, but activities with significant consequences require controls, traceability and oversight proportionate to the risk.
Security, pilot and growth
Every new workflow changes who can access data and how it is retained. Least-privilege permissions, backups, audit trails, access management and data protection must be defined during design. Privacy by design also requires collecting only the information needed and establishing its purpose and retention period.
A sustainable approach is to select one process, run a controlled pilot and compare the results with the baseline. The solution is extended only after validation. Lead time, errors, manual work, deadlines met and actual usage are more meaningful indicators than the number of features installed.