1. Why notification is a matter of hours
When something serious happens, the first question is not technical but organisational: who knows, who decides and who writes to the authority. NIS2 sets an early warning within 24 hours and an incident notification within 72 hours from the moment the organisation becomes aware of it. The clock runs from awareness, not from the end of the investigation, and this changes how the procedure must be designed.
A procedure designed for normal working days often fails precisely in the situations where it is needed. An attack at the weekend, an overnight outage or the absence of the main contact put to the test the decisions taken at the design stage. For this reason notification must be prepared as a process with roles, on-call cover and templates already in place, not improvised in the middle of an emergency. For this reason, too, the procedure should be thought of as an on-call rota, with people and tools ready.
2. Establishing whether the incident is significant
Not every incident must be notified. The obligation concerns significant incidents, meaning those with material effects on the service or on people. The assessment must be made quickly, using written criteria, and must be documented even when the conclusion is that the event is not significant. The table below proposes general criteria to adapt to your organisation and to check against the text of the legislation.
| Criterion | Question to ask | Generic example |
|---|---|---|
| Service disruption | Is the service delivered to customers down or materially degraded? | A booking system does not respond for the whole day |
| Loss or compromise of data | Has sensitive or critical data been exfiltrated, encrypted or altered? | A file server is encrypted by malware |
| Financial impact | Does the event cause significant losses or block revenue-generating activity? | A production stoppage that blocks deliveries |
| Impact on others | Could the event affect customers, suppliers or other parties? | Customer data sent to the wrong recipients |
| Suspected malicious action | Are there signs of an intentional act or unauthorised access? | Unusual access using an employee's credentials |
3. The sequence: early warning, notification, final report
Communication takes place in several stages. The early warning is an initial, quick and concise message stating that a significant or suspected significant incident exists and that it may have cross-border or widespread effects. The incident notification follows, within 72 hours, and contains a first, more structured assessment: severity, impact, available indicators and measures already taken. The final report closes the process with a full description, the causes and the planned remedial action.
The timelines for the final report and for any interim updates must be checked against the applicable text and the authority's guidance. Do not set in the internal procedure a date that you have not verified. You can instead establish that the final report starts immediately after the incident is closed, with a responsible person and a template already prepared. Whoever signs the report must be able to explain every statement with a document or a technical piece of evidence.
4. What each communication must contain
Each communication has a different purpose, and its content must be consistent with the stage. The early warning does not require a full analysis, but it must be precise about the existence of the event. The notification must be richer. The final report must explain what happened, why and what was done about it. In all cases, use simple and verifiable language, and clearly distinguish established facts from hypotheses.
- Early warning: date and time of awareness, short description, suspicion of malicious action or wide impact, name of the contact person.
- Notification: initial assessment of severity, services and systems involved, technical indicators available, containment measures already taken.
- Final report: sequence of events, root cause, actual impact, corrective and preventive measures, any effects on other entities.
- In all stages: references to supporting documents and a clear statement of what is still being verified.
5. Who must send it, and with what mandate
The communication must be sent by a person with a formal mandate, usually the designated contact, with a deputy of equal authority for cases where the first is unavailable. Management must know who is authorised to send and sign, because urgent decisions cannot wait for a meeting. Involve legal counsel when the incident affects personal data or contracts, and the data protection officer when the event concerns personal information.
6. The operational procedure, step by step
A useful procedure is short and readable in a few minutes. It must state who does what, in which order and with which templates. Each step leaves a trace with time and owner, because the chronology is one of the elements the authority assesses. Test the procedure at least once with a tabletop exercise, simulating an incident at night. The record of the exercise, with date, participants and gaps identified, is a useful document to keep.
- Detect the event through monitoring, user reports or supplier alerts, and record the time of first awareness.
- Activate the on-call contact and open an incident record with a unique identifier.
- Quickly assess significance using the written criteria and note the decision with its reasons.
- If the incident is significant, prepare the early warning from the template and send it within the required timeframe.
- Contain the event, preserve evidence and collect technical indicators before changing systems.
- Prepare the notification with the initial assessment, updating it if new elements emerge.
- Restore services following the continuity plan and document every action.
- Write the final report with causes, impact and corrective actions, and assign each action an owner and an internal deadline.
7. The link with personal data breach notification
An incident may also involve personal data, in which case another obligation under the GDPR is triggered: notification of the breach to the data protection authority, generally within 72 hours of becoming aware of the breach. The two communications have different recipients, contents and legal bases, but they arise from the same event and often share the same timelines. The supplier that runs the service should also know this coordination point.
For this reason the procedure must provide a coordination point at which the NIS2 contact and the data protection officer assess the event together. Avoid contradictory communications: the facts described must be the same, even if the form changes. Check with the legal adviser how to combine the two routes in your case. Document the decisions taken in this exchange, because they will be the element that explains why a communication was sent with a given content.
8. Common mistakes and related guides
The most frequent mistakes are almost always organisational. Waiting for the end of the investigation before starting to notify, entrusting the decision to a person who is not reachable, writing a communication without having verified the facts, or forgetting that the supplier has its own communication obligations towards you. The contract with the supplier must provide notice periods compatible with yours.
To explore the scope of the legislation, read the guide on classifying entities. For the general structure of the work, read the guide on the compliance path. For the content of supplier contracts, read the one on the supply chain. If you need to prepare a procedure from scratch, start from the gap analysis checklist and then return here for the notification sequence.
9. How to start
The first action this week is to decide who the notification contact is and who their deputy is, with a written mandate. Immediately afterwards, check with your IT provider who monitors systems outside office hours and how the contact is alerted in the event of an incident. These are two answers that can be obtained in a short meeting and that immediately improve the quality of your response.
Then prepare three templates: the early warning, the notification and the final report. They do not need to be perfect. They need defined fields and a completed example. Finally, organise a tabletop exercise within the next few weeks, and note what did not work in order to correct it before it happens for real. Write the templates in clear language, with mandatory fields highlighted, so that they can be completed even by someone who has never used them.
Frequently Asked Questions
From when do the 24 hours of the early warning run?
The clock starts when the organisation becomes aware of a significant incident, or has a reasonable suspicion that it is one. For this reason it is important to record the time of first awareness precisely, because it will be the reference point for assessing timeliness. Do not wait for complete confirmation of the incident before starting to assess notification.
Do I have to notify an incident I have already resolved?
If the incident is significant, notification is still due even if the problem was resolved quickly. The assessment is based on the facts, not on the outcome. Keep the documentation of the assessment in any case, even when you conclude that the event does not meet the criteria, with the reasons and the name of the person who approved it.
Can my IT provider send the notification on my behalf?
A provider can prepare the technical content and support the drafting, but the communication must be sent by a person authorised by the organisation, and responsibility remains with it. Define in the contract who does what, with which notice periods and which templates, and make sure the contact can always approve the text before it is sent.
What happens if I cannot provide all the information within the deadlines?
The initial notification can be completed with later information, provided that the first communication is timely and honest about what is still being verified. Do not delay sending while waiting for complete details. State the gaps clearly and the commitment to update, then honour that commitment with subsequent updates.
Do I have to notify if the attack was blocked by the firewall?
A blocked attempt is not always a significant incident. It depends on what was targeted, the intent and the potential effects. An automated attack rejected by standard controls is often an event to record rather than to notify. Document the assessment using the written criteria and, if you have doubts, ask an adviser for an opinion.
Do I need a separate register for incidents?
An incident register is useful in every case, including for non-significant incidents, because it shows the trend of events and the quality of the response. It must contain the date, description, assessment, notification decision, actions taken and closure. A well-kept register is also the basis of the final report and of the authority's checks.
Who inside the organisation should be told first?
The notification contact, for the technical assessment and the decision. Immediately afterwards management, which must be informed when the incident is potentially significant, and the data protection officer if personal data is involved. The procedure must set out these steps in order, with contact details that can be reached outside office hours.