GUIDE · DORA

DORA exit strategy for critical ICT services: how to write it

After this guide you will know how to write an exit strategy for critical ICT services, meaning the plan that allows you to leave a supplier without interrupting the business, and how to check that the plan is workable before it is needed. The work requires collaboration between IT, purchasing, compliance and the functions that use the service.

Updated october 2026
8 min read
Altovar Technical Editorial Team

1. Why the exit strategy matters

An exit strategy answers a simple question: if we had to leave this supplier tomorrow, because of a dispute, a failure or a strategic decision, what would happen? For critical ICT services the answer must be written down and verified. Without a plan, dependence on a supplier becomes a constraint that only emerges at the worst moment, when the options are already limited.

DORA requires exit to be considered in two modes. The first is orderly: a planned transition to another supplier or to an in-house solution. The second is under stress: a supplier that can no longer deliver the service, either suddenly or gradually. The plan must cover both, because timeframes and risks are very different. They should be considered separately, each with its own timing.

2. Prerequisites: what you need before writing

The plan depends on three basic elements. The first is a map of critical or important services, with the business processes they support. The second is the contract, with its clauses on term, termination, return of data and assistance during transition. The third is technical knowledge of the service: where the data is held, in which format and with which dependencies.

If any of these elements is missing, the plan cannot be complete. Involve the owner of the process that uses the service, because only they know how long it can be without it, and the IT manager, who knows what it really means to move data. Your legal adviser is needed to verify that the termination and data return clauses are actually enforceable.

3. How to write the plan, step by step

The plan should be built in sequence, because each step produces a document that the next one uses. Do not try to write everything at once: start with the most critical services and apply the same template to the others. The end result is an exit strategy for each critical service, with a concise summary sheet and the technical annexes required. A long document is not necessary.

Write the plan in operational language. A sentence such as “ensure continuity” says nothing to someone who has to act at night. A sentence such as “export the case archive in the agreed format and load it onto the alternative” indicates an action that someone can carry out and verify. Avoid generic phrasing: each step should state who does what, with which tool and within what time, even if the timings are estimates.

  1. Describe the service, the process it supports and the impact of an interruption of varying duration, using the assessments already made by process owners.
  2. List the dependencies: connected systems, integrations, data, users and other suppliers that rely on the service.
  3. Identify possible alternatives, internal or from other suppliers, with the minimum requirements they must meet.
  4. Define the activation conditions for the exit, meaning the events and thresholds that trigger the plan, and who decides.
  5. Describe the transition steps: data extraction, formats, migration, functional testing and return.
  6. Assign responsibilities for each step to each function, with contacts and timeframes to be agreed.
  7. Verify with the supplier the clauses the plan genuinely needs, such as transition assistance and data return.

4. Common mistakes

Weak exit strategies tend to fail for the same reasons. A quick test is to ask who can activate the plan at night, with which credentials and where they find the instructions. If the answer requires more than one phone call, the plan needs simplifying. The plan must be accessible even when the main system is not, because that is precisely when it will be consulted.

  • Writing an exit clause and considering the problem solved, without checking that the transition is workable.
  • Not keeping a copy of the data in a readable format that other systems can use.
  • Ignoring third-level dependencies, meaning the suppliers of your suppliers.
  • Preparing the alternative only on paper, with no functional testing.
  • Leaving the plan in the hands of a single person who then changes role.

5. How to check that the plan is workable

A plan that has not been tested remains a statement of intent. Verification can be documentary, through a cross-review of contracts and dependencies, or practical, by exporting a sample of data and loading it into a test environment. The second is more demanding but far more convincing. Both types of verification should be planned, with different objectives and timescales.

Record the results: which data was exported, how long it took, which errors emerged and how they were corrected. This evidence is what you show in the event of an inspection, and the material for updating the plan. Repeat the verification when the service, the contract or the alternative supplier changes significantly. Also keep previous versions of the plan.

6. What to keep up to date

The plan ages. Systems change, suppliers change, people change. A plan written some time ago may name an alternative that no longer exists or a contact who has left the organisation. Set a periodic review and a trigger for extraordinary reviews, such as a contract renewal, an incident or a change of supplier. Review it at least once a year, without waiting for an event.

The supplier register and the exit strategy must talk to each other: every critical service in the register must have a linked plan, and every plan must point to existing register entries. This link is easy to lose, and an inspection may start from a random service to test it. Keep this link in the checklist for the annual review, alongside the other contract controls, and spend a few minutes on each.

7. Points to clarify with your adviser

Some parts of the plan depend on the contract and on the applicable regulatory framework. Ask your adviser whether the clauses on data return, transition assistance and termination are adequate, and whether the plan meets the supervisory requirements for your entity. Also ask which evidence the competent authority expects to see, and in what form. Always ask in writing.

Finally, check that the alternative supplier, if already identified, does not belong to the same dependency chain as the current supplier. Two suppliers that use the same sub-supplier do not offer a genuine alternative, and this is a point to clarify in writing with both. Verify that the shared sub-supplier is not critical to the alternative as well, and obtain written confirmation from both suppliers.

8. How to get started

This week, choose the most important critical service and write the question from which everything starts: if this supplier were no longer available tomorrow, who would notice first and what would they do? The answer, even if incomplete, reveals the dependencies missing from the plan. Do not try to complete it in a single meeting. The answer does not need to be complete; it needs to be honest.

Then collect the contract for that service and check the termination and data return clauses. Bring the list of questions to your adviser and to the supplier. The plan starts from there: a first document for a single service, tested on a small scale, is worth far more than a general plan for all services. Bring a proposed timetable, with dates and owners.

Frequently Asked Questions

Does every ICT supplier require an exit strategy?

No. The plan is required for critical or important services, that is, those whose interruption would compromise the business or supervisory requirements. Other suppliers must still be registered and managed, but with a lighter level of detail. The distinction should be justified within the risk management framework and verified with your adviser.

Can an exit plan include bringing the service back in-house?

Yes, the exit can be towards an in-house solution, if the organisation has the skills and capacity to run the service. The plan should assess the costs, timeframes and risks of this choice, without assuming it is simpler. An in-house solution often requires a transition period that needs to be planned in detail.

How do I handle an exit if the supplier suddenly fails?

A plan under stress cannot rely on the supplier's cooperation. You must already have accessible copies of the data, system documentation, alternative contacts and a pre-qualified alternative. Data return and escrow clauses must be verified in advance, because it is too late afterwards.

How detailed should the plan be?

Detailed enough to be carried out by a competent person who did not take part in writing it. The level of detail grows with the criticality of the service. A plan that is too generic is of no use, and one that is too long will not be read. A good test is whether the actions are clear, assigned and linked to real data and contacts.

Should the plan be shared with the supplier?

Partly. Some information, such as export formats and assistance timeframes, should be agreed with the supplier and included in the contract. Other parts, such as the alternatives evaluated, remain internal and confidential. Ask your adviser which elements should be shared and in what form.

About this content

General informational guide. Regulations and technical requirements change over time: always check the official sources and assess your specific case with a qualified technical or legal advisor.

GET IN TOUCH

Talk to our team.

We'll show how Altovar integrates into your infrastructure. No commitment, no sales pitch. Just a technical conversation.

BOOK A DEMO →