1. What a business continuity plan is
A business continuity plan, known as BCP, describes how the company keeps carrying out its essential activities when a serious event interrupts normal operations. A fire, a ransomware attack, a prolonged connectivity failure, the sudden absence of a key person or the failure of a supplier are typical examples. The plan does not concern computers alone: it covers people, sites, processes, suppliers and communications with customers and staff.
Confusing the BCP with the disaster recovery plan is the most frequent error. Disaster recovery concerns the restoration of the IT infrastructure, meaning servers, data and applications. The BCP comes first and contains it: it establishes which activities matter most, the order in which they return and which alternatives are used in the meantime, even when the systems are not there. A good plan answers a concrete question: what do we do on Monday morning if the management software will not open?
2. Why a mid-sized company needs one
In small companies, continuity often depends on the memory of a few people. In mid-sized companies complexity grows: more sites, more suppliers, more interconnected systems, more customers expecting prompt answers. When an event hits one of these points, without a plan decisions are made quickly, inconsistently and often wrongly, because each person defends the part they know best and nobody sees the whole picture.
The plan has value beyond the emergency. It forces you to describe processes in writing, identify hidden dependencies and clarify who decides. Customers and partners increasingly ask to see that this work exists, and anyone wanting a formal reference can look to the ISO 22301 standard, covered in a dedicated guide. In every case the content must withstand the test of a real event, not just a paper check carried out once.
3. Scope: processes, systems and people
The first step is to decide what is included. Start from processes, not technologies. For each business function ask which activity cannot stop for more than a few hours, which can wait a few days and which can be suspended without serious consequences. This distinction guides the rest of the plan and prevents spending on protecting things of little importance while leaving uncovered what really matters.
Then link each process to the systems, data, people and suppliers that support it. An order depends on the management software, email, the label printing system, the courier and the bank account used for payments. Writing out this chain reveals dependencies that nobody remembered. You do not need a perfect inventory on the first pass: you need a map complete enough to set priorities. The detail grows in later revisions, once the owners have verified the starting data. Typically the scope includes:
- Processes that generate revenue or meet legal and contractual deadlines
- Systems, applications and data that support them
- People with unique skills and their substitutes
- Physical sites and possible alternative spaces
- Critical suppliers and their ability to keep serving you
4. Impact analysis: the starting point
The business impact analysis, known as BIA, is the stage where you measure what each interruption costs. For every critical process the consequences over time are assessed: financial loss, harm to customers, obligations towards third parties and impact on reputation. The result indicates two values: the maximum tolerable interruption time and the amount of data that can be lost without serious harm. The first is the recovery time objective, the second the recovery point objective.
The BIA must be carried out with the people who truly know the process, that is, operational managers, not only the IT department. Qualitative estimates are acceptable provided they are reasoned and approved. Do not invent precise figures to appear rigorous: state your assumptions and revise them after every test. The step-by-step method is covered in the dedicated BIA guide, which you should read straight after this one.
5. Scenarios and recovery strategies
The plan must provide for concrete scenarios, not a generic 'in case of emergency'. A few representative ones are enough: loss of a site, unavailability of information systems for technical reasons or due to a cyber attack, interruption of connectivity, prolonged absence of key staff, failure of a critical supplier. For each, state who declares the state of emergency, which processes are activated and in what order, so that the response does not depend on improvisation.
For each critical process choose a recovery strategy. This may be working in manual mode, moving to an alternative site, using a standby system or restoring from a backup copy. Not all alternatives have the same cost or the same activation time. The comparison between a hot, warm or cold disaster recovery site is a specific topic, covered in a dedicated guide.
6. External dependencies and degraded mode
Many interruptions originate outside the company. Connectivity, electricity, telephony, cloud software providers, couriers, banks and consultants are dependencies that the plan must name. For each, ask the supplier whether it has its own continuity plan, what recovery times it commits to and how it will alert you in the event of a problem. The answers should be written into the contracts, not left only in the internal document, because in an emergency what is written and signed is what counts.
Degraded mode is the ability to work partly even without systems. For each critical process write what can be done on paper or with local tools, who authorises it, how operations carried out during the emergency are recorded and how they are entered into the systems when these become available again. Without these rules every emergency becomes an improvised discussion, and manual operations risk being lost along the way.
7. Testing, updating and ownership of the document
An untested plan is a hypothesis. Schedule periodic exercises, starting with a paper-based walkthrough in which the crisis group works through a scenario and discusses it. Then move on to an operational simulation and, when maturity allows, to a real recovery test. Every exercise must produce a list of corrections, each with an owner and a deadline. Without this step the exercise becomes a pleasant memory and nothing changes in the document.
The document has an owner, usually management or an appointed manager, and must be reviewed at least once a year and whenever systems, sites, suppliers or the organisation change. Keep a single approved version, with a date and a change log. Printed and offline copies must be updated together with the digital version, because in an emergency nobody checks which is the most recent, and an old copy can lead to wrong decisions.
8. The cluster's deep dives
This guide is the overview of a group of continuity deep dives. You do not need to read them all at once, but it helps to know which questions they answer. Each topic solves part of the problem and refers to the others, because the BIA feeds the strategy, the strategy feeds the plans and the plans must be tested. If you are drafting the first version of the plan, these are the topics you will encounter, to be tackled in the order you need them.
If you are starting from scratch, read the BIA first, then the disaster recovery plan, then the dependency mapping: these are the three documents your plan will use directly. The other guides are useful when you reach the technical choices, the tests and the compliance checks. It is best to read them once you have a draft to compare against, because the questions become much more concrete and the answers more useful.
- Business impact analysis: priorities and tolerable times
- IT disaster recovery plan: structure of the technical document
- Hot, warm and cold recovery sites: how to choose the model
- Failover and failback: switching to the standby system and back
- Disaster recovery testing: levels of testing
- Critical dependencies: connectivity, energy and suppliers
- Operational runbooks for the critical moment
- ISO 22301 for those who want a formal management system
9. How to start
This week, do just one thing: choose three processes that, if they stopped, would cause immediate damage to customers or payments. For each, write on a sheet who runs it, which systems and suppliers it uses and how long it could stay stopped before causing serious problems. This draft is the first input to the BIA and will immediately reveal the biggest gaps, especially where only one person knows the procedure.
Then organise a short meeting with management and the managers of those processes, to validate the draft and appoint who coordinates the plan. Bring one closing question for your IT supplier: which parts of continuity do they manage for you and which remain your responsibility. The answer shows where the plan must be written in-house and where it can be delegated, and it should be recorded in writing in the document.
Frequently Asked Questions
What is the difference between a business continuity plan and a disaster recovery plan?
The business continuity plan (BCP) covers the whole of the company's activity: people, sites, processes, suppliers and communications. The disaster recovery plan is its technical part and concerns restoring servers, data and applications. The BCP sets priorities and decides what to do when systems are unavailable; disaster recovery describes how to bring them back into operation. The BCP is usually written first, because it provides the restoration order that disaster recovery must respect.
Who should write the business continuity plan?
The plan should be written together. Management approves it and assigns responsibility, process owners provide the operational content, and IT describes the systems and restorations. An external consultant can facilitate the work and check its consistency. Do not delegate everything to the IT supplier: they know the technology, but do not always know which of the company's activities matter more than others.
How long should a continuity plan be for an SME?
Length matters less than clarity. A useful document in an SME is usually an orderly collection: a general section, a list of critical processes with their strategies, a directory of roles and contacts, and the essential operating procedures. If the document is hard to consult during an emergency, it is too long, however complete it may be on paper.
How often should the continuity plan be updated?
An annual review is a reasonable cadence for most companies, but the plan should also be updated after every significant change: a new site, a new management system, the change of a critical supplier, a reorganisation of roles or a real incident. Also check that contacts are still valid, because they are the part that ages most quickly.
Is a continuity plan required by law?
It depends on the sector, the size and the type of activity. Some regulations, such as the NIS2 Directive as transposed into Italian law by Legislative Decree 138/2024, require cybersecurity risk management measures that include continuity, and regulated sectors may have specific obligations. For a mid-sized company the answer should be verified with a consultant or the competent sector authority, without assuming either that it is mandatory or that it is not.
Is ISO 22301 certification needed to have a good plan?
No. An effective plan can exist without any certification. The ISO 22301 standard defines a business continuity management system, with review and improvement cycles, and it is mainly useful if you want to demonstrate this formally to customers or partners. Before considering certification, it is better to have a written plan that has been tested at least as an exercise.
What should I ask my IT supplier about continuity?
Ask which continuity services are included and which are not, how restores are tested and how often, what the supplier's own dependencies are, how they will alert you in the event of an incident and who decides in case of conflict. Also check whether responsibility for data and restores is described in the contract, because that is where what really happens in an emergency is established.