Start with the service and plausible harm
A hospital needs continuity of care and trustworthy clinical information; a drinking water operator may prioritise safe treatment and pumping; a railway infrastructure manager may focus on operational control and traffic availability. These are different outcomes even when all entities use networked information systems. Identify the regulated service, major dependencies and consequences of interruption before ranking security projects. The Annex I sector label is the beginning of the assessment, not its conclusion.
Apply proportionality without weakening the statutory duty
Article 21 directs covered entities to take appropriate and proportionate technical, operational and organisational measures. Consider exposure, size, implementation cost, and the likelihood and severity of incidents, including societal and economic impacts. Proportionality does not mean choosing the cheapest control regardless of risk. A modest-sized entity operating a sensitive service may require stronger safeguards around critical functions than a larger organisation with less consequential exposures.
Different sectors need different resilience tests
A data centre may test loss of power, cooling or upstream connectivity. A rail operator may rehearse loss of dispatch visibility. A healthcare provider may need a safe downtime procedure for electronic prescribing, while a space ground-station operator may test command-link recovery and compromised access. Each test should have defined objectives, operational safety protections, observations and corrective actions. Generic tabletop exercises without an identifiable service outcome are weak evidence of resilience.
Treat supplier and operational technology risk precisely
Supply-chain security is a shared Article 21 obligation, but dependency patterns differ. Hospitals may rely on clinical technology vendors, water operators on industrial control maintenance, and MSPs on remote administration platforms serving many customers. Identify which suppliers can affect essential service continuity or system integrity. For industrial and safety-related equipment, select vulnerability management and change procedures suited to operational constraints rather than automatically applying ordinary office IT maintenance practices.
Translate sector risk into management decisions
Management bodies must approve and oversee the measures adopted for Article 21 under Article 20. Present major service-risk scenarios, control gaps, recovery-test findings, owners and remediation deadlines. Use sector-specific indicators rather than assuming one dashboard suits all entities. Keep the legal classification, risk rationale, competent-authority contacts and incident notification criteria current. Review priorities after material incidents, new technologies or changes in service delivery.
Points of clarification
Does NIS2 require identical controls in every Annex I sector?
No. Article 21 requires appropriate and proportionate measures responding to the entity's risks and circumstances.
Should risk priorities depend only on company size?
No. Exposure, incident likelihood and severity, operational impact and other factors also matter.
Why is sector-specific recovery testing important?
It tests whether the organisation can continue or restore its regulated service under realistic disruptions.