1. What DORA requires, and why it is not a box-ticking exercise
DORA, short for Digital Operational Resilience Act, is the European Regulation that requires financial entities to demonstrate that they can withstand, respond to and recover from disruptions and incidents linked to information and communication technologies, in short ICT. The Regulation is not satisfied with a well-written security document. It requires processes that work over time, with clear responsibilities and evidence that can be shown to supervisors.
For this reason it is best to think in terms of a programme. A programme brings together risk management, incident management, testing, oversight of providers and information sharing, and evolves them together with the organisation. Those who treat DORA as a list to tick off once end up with documents that nobody updates and with systems that change far faster than the paperwork. At the first inspection, the result is a gap between what is written and what actually happens.
2. Before you start: scope, roles and starting data
Before writing any plan you need to understand which entities the Regulation applies to and with what obligations. The answer depends on the type of authorisation and the activity carried out. The Regulation provides a principle of proportionality, meaning obligations calibrated to size, complexity and risk, but this does not exempt anyone from governing ICT risk. Check with the compliance function or with a legal adviser whether your organisation falls within scope and which version of the obligations applies to you.
You also need three starting elements: a list of ICT systems and the services that support them, a map of who is responsible for each system, and the existing documentation on security, continuity and incident management. They do not need to be perfect. They need to exist and be comparable with what the Regulation requires. An initial survey of this kind usually takes a few weeks of shared work between IT, compliance and procurement, and its output becomes the starting point for everything else.
3. The five pillars and where to read more about each
DORA is structured around five areas. The programme must cover all of them, but not all of them start from the same level of maturity: the order of work depends on which functions are already robust in your organisation. The table summarises each area, what the Regulation asks for in practice and the topic to explore in a dedicated guide. You do not need to read them all at once: it is better to start with the area where you are weakest. Compare each row with what already exists.
TLPT, threat-led penetration testing, is the most advanced form of testing. It does not apply to all entities in the same way: any obligation depends on designations by the competent authority, so it must be confirmed case by case. Before reaching it, it is important to have consolidated the basic tests, such as restoration checks and incident simulations. It is worth understanding these before planning anything further.
| Area | What it requires in practice | Topic to explore |
|---|---|---|
| ICT risk management | A documented framework, approved by the management body, describing how to identify, assess and control ICT risks | ICT risk management framework: structure and content |
| Incident management | Criteria for classifying incidents and procedures for reporting serious ones to the authorities | Classifying and reporting serious ICT incidents |
| Resilience testing | A testing programme ranging from basic checks to advanced tests for the most exposed entities | DORA resilience testing: from basic tests to TLPT |
| ICT third-party providers | A register of provider contracts, assessment of concentration risks and exit strategies for critical services | DORA information register: what the Regulation requires |
| Information sharing | Rules for exchanging threat information between entities, with confidentiality respected | Threat information sharing under DORA |
4. The steps of the programme, in order
The route below is a sequence of work, not a hierarchy of importance. Each step produces a result that the next one needs, so it is best to respect the order for the first stages. Some activities, such as reviewing contracts and drafting incident procedures, can proceed in parallel when resources allow. Do not start a step if the previous one has not produced a verifiable result.
These steps do not close into a single project. After the first version, the programme enters a maintenance phase: every significant system change, every new provider and every material incident requires a targeted update of the parts involved. It is maintenance, more than the first draft, that determines whether the programme holds up over time. Those who neglect it discover the gaps only during an inspection.
- Map the ICT services and the providers that support them, indicating for each service the system, the data and the business process involved.
- Identify the critical or important services, meaning those whose interruption would compromise the business or compliance with supervisory requirements.
- Write the ICT risk management framework and have it approved by the management body in a minuted discussion.
- Build the provider register and link every ICT contract to the criticality of the service it supports.
- Define the incident classification criteria and reporting procedures, with owners and ready-to-use templates.
- Plan testing starting from basic checks, such as backup restores and incident simulations, and decide when to move to more advanced levels.
- Write exit strategies for critical services and verify that transfer to another provider is actually feasible.
- Set a periodic review of the programme and a channel for tracking regulatory updates and the authority's guidance.
5. Governance: who decides and who executes
DORA places responsibility on the management body, not only on the IT function. In practice the management body must approve the ICT risk strategy, oversee its implementation and have the skills needed to understand it. A programme that senior management does not know about stops at the first difficult budget request, because nobody has the authority to back it. For this reason, the first request to the top is the mandate for whoever leads the work.
The programme needs a point of contact with real authority and direct access to senior management. Alongside that person, a responsibility matrix is needed indicating, for each area, who writes, who approves and who executes. The security lead, the systems lead, the compliance function and whoever manages the purchasing of IT services must all read the same matrix. If each function believes that the check belongs to someone else, none of them carries it out.
6. Common mistakes to avoid
The most common mistakes are predictable, and for that reason they can be avoided. Check your organisation against this list before presenting the programme to the management body. In particular, those approaching this work for the first time tend to repeat very similar mistakes, so it is worth knowing them before writing a single line of the document. The first mistake to avoid is relying on standard templates, which rarely describe real systems.
The most effective remedy is a review of the programme at least once a year with the owners of each area. The review compares what is written with what actually happens, using incident logs, test minutes and contract changes. The differences that emerge become a remediation plan, with owners and clear deadlines, and the plan is reported back to the management body.
- Treating the ICT risk management framework as a document to deliver once, without periodic reviews.
- Building the provider register without linking it to the critical services and business processes it supports.
- Planning sophisticated tests before verifying that backup restores really work.
- Involving providers only in formal tests, discovering hidden dependencies only during a real outage.
- Writing exit strategies as theoretical clauses, without ever verifying that a move to another provider is feasible.
- Forgetting that the programme must be updated after every significant change of system or provider.
7. How to verify that the programme works
A programme is judged by evidence, not by documents. For each area, ask for a concrete proof: a restore performed and recorded with its outcome, a simulated incident with the times noted, a contract showing the criticality of the service, or minutes in which the board discussed and approved the risk framework. If a proof is missing, record it as a gap with an owner and a date.
Record what did not work, too. A failed test, documented together with the corrective action taken and the subsequent verification, demonstrates maturity more than a test never carried out. Results must be reported to the management body in a simple summary: what was tested, what emerged, what changes in the plan and who is responsible. A documented outcome of this kind is the most convincing evidence of a living programme.
8. Which deeper guides to read next
This guide provides the overview. The topics that merit a separate deep dive are the content of the ICT risk management framework, the information register on provider contracts, the criteria for classifying and reporting serious incidents, the testing programme up to TLPT, the exit strategy for critical services, proportionality for smaller entities, threat information sharing, concentration risk on a small number of providers and preparation for an inspection.
If you read only one deeper guide after this one, choose the one on the ICT risk management framework. It sets the language, responsibilities and criteria on which the other areas rely. Right after it, read the one on the provider register, because collecting contract data is often the most demanding piece of work and determines how solid the other areas can be.
9. How to start: the first week
The first useful step is small and can be taken this week. Gather the heads of IT, the compliance function and whoever manages contracts with technology providers. Ask each of them to bring the list of systems and external services they know of, even if incomplete. Do not seek agreement on everything: the aim is to understand the real scope. Note the uncertain answers.
With that list, prepare a first map of ICT services and mark those which, if they stopped, would halt the business or put customers in difficulty. That map is the point from which to start the risk framework, the provider register and the tests. Present it to the management body as a proposed work plan, with the calendar of the next steps.
Frequently Asked Questions
Does DORA apply to a small financial services company?
It depends on the type of authorisation and the activity carried out, not only on size. The Regulation applies the principle of proportionality, so obligations are calibrated to complexity and risk, but it does not exempt anyone from managing ICT risk. Check with the compliance function or an adviser whether you fall within the scope.
Where do I start if I do not yet have any documents?
You do not need to start with writing. Begin with the map of ICT services and providers, because without it every document would be a guess. Once the map is ready, the risk management framework becomes easier to write, because it describes systems that really exist and responsibilities already identified.
Does my cloud or hosting provider handle DORA for me?
No. The provider can supply documentation, information for the register and contractual clauses, but responsibility for risk remains with the financial entity. Ask the provider which data it can produce for the register, how it handles incidents that affect it and what conditions apply at the end of the contract, and get everything in writing.
How long does a complete programme take?
It depends on the starting point, the complexity of the systems and the number of providers involved. An organisation with already mature documentation can complete the first version more quickly than one starting from scratch. In any case the programme is never closed: it must be reviewed after every significant change and periodically.
Do I need specific software to manage the programme?
No particular product is mandatory. A set of shared spreadsheets may be enough for a small organisation, while more complex organisations often use dedicated tools for the provider register and incident management. What matters is that the data is up to date, accessible to those who need it and verifiable in the event of an inspection.
What happens if an incident involves a provider?
The provider must inform you under the terms of the contract, and you assess the incident using the criteria you defined to classify it. If it falls among the serious ones, reporting must be done within the timeframes set by the Regulation and the related technical standards, which you should verify with your adviser. This is why the notification clauses must be written before the incident, not during it.