1. What a business impact analysis is and what it is for
A business impact analysis, abbreviated to BIA, measures what happens to the company when a process stops, and for how long. It does not start with servers: it starts with the activities that create value, the deadlines that must be met and the commitments to customers and suppliers. The result is a ranking of processes, showing which must be restored first and which can wait, and what consequences to expect in the meantime.
The BIA is the foundation of the business continuity plan. Without it the plan decides priorities on gut feeling, and often rewards the most visible system rather than the most important one. With the BIA you can justify to management investments in backups, alternative sites or manual procedures, because every choice is linked to a described and approved impact. This makes the discussion concrete, because it deals in consequences rather than opinions.
2. Who takes part and when to do it
The BIA should be carried out with process owners, meaning the people who know how a function really works: administration, sales, logistics, production, customer service, each for their own area. The IT manager takes part to clarify which systems support each activity, but cannot estimate the business impact alone. Management approves the result and decides priorities when two processes conflict, because choosing between two urgent needs is a business decision.
It is best done when the plan is first launched and then every time the organisation changes significantly: a new management system, an additional site, a supplier becoming critical, an acquisition. You do not need to wait for a perfect reorganisation. A well-done BIA on a limited scope is more useful than a complete one that is years old, because decisions should rest on today's data, not on an organisation that no longer exists.
3. Step one: inventory of processes
Start from a list of processes, not a list of systems. For each function, collect the main activities and note some basic data for each. This initial table becomes the basis for all the work that follows and must be validated with the people who perform it every day, because real processes often differ from those described in organisation charts. Use a simple sheet, with one row per process and columns for the later stages.
Do not aim for completeness on the first pass. A reduced but verified inventory is worth more than a broad document nobody has read. When processes emerge that nobody knows who manages, flag them as a gap to close, not as a detail to ignore. You will find that some processes overlap or have different names in different departments: merge them only after verifying their real content. For each process record at least these elements:
- Process name and the function that manages it
- Main owner and their deputy
- Systems, data and applications it uses
- Suppliers and external services involved
- Fixed deadlines, such as payments, filings or deliveries
- Sites and workstations where it takes place
4. Step two: measuring the impact over time
For each process, ask how the impact changes as time passes. An outage of a few hours may be almost painless, whereas after a few days problems start with customers and suppliers, and after a few weeks fines or lost contracts may follow. Always reason in intervals and with concrete questions, not with absolute figures that nobody can verify. Ask the owners to indicate the moment when the problem becomes visible outside the company, because that is often where the impact truly grows. The table summarises the dimensions to examine.
Record the assumptions you use. If you believe a delay of a certain length is tolerable, write down why and who confirmed it. When the process owner disagrees, record the disagreement: it is often the point where the plan must be revisited, and it should not be resolved by averaging the two opinions. An open discussion on this point is often more useful than a hasty agreement.
| Dimension | Question to ask | Who answers |
|---|---|---|
| Financial | How much is lost if the activity is stopped, in revenue or extra costs? | Management and finance |
| Customers | Which customers notice, and after how long? | Sales and customer service |
| Obligations | Which legal or contractual deadlines must not be missed? | Legal or compliance manager |
| Reputation | What is said externally if the service stops? | Management and communications |
| People and data | Are there risks to security or to special category data? | Security manager |
5. Step three: recovery times and tolerable data loss
Two objectives follow from the impact measurement. The first is the recovery time objective (RTO): the maximum time within which a process must be operational again before the impact becomes unacceptable. The second is the recovery point objective (RPO): how much recent data can be lost without serious harm, that is, how far back the copy used for restoration can go.
These objectives are requirements, not promises. A very short RTO requires more expensive solutions, such as standby systems already in place, and a very low RPO requires more frequent copies or data replication. Ask your supplier what times it can guarantee for each service, and verify that they match those you have approved internally. Any discrepancy must be resolved before an emergency, not during it, when nobody has time to negotiate.
6. Step four: dependencies on systems and suppliers
A process has a realistic recovery objective only if its dependencies do too. If the process needs a management system that in turn depends on an authentication server, an internet connection and a payment provider, the real time is that of the slowest link in the chain to reactivate. The BIA must therefore connect each process to its systems and external services, and ask each supplier what recovery objectives it commits to. This step requires listening to whoever runs the service, not just reading the documentation.
This step often reveals bottlenecks: a single shared administrative account, a document that exists only on one computer, a signature that depends on one person. Note these elements too, because they are not technical details but real risks to the business. The guide on critical dependencies explains how to map them systematically, with a repeatable method. Each new map should follow the same scheme, so that comparing them over time is straightforward.
7. Using the results and common mistakes
The BIA results serve to order recovery, choose the strategies for each process and decide where to invest. A ranking approved by management becomes the basis for the IT disaster recovery plan, the choice of recovery site and the operational runbooks. If the BIA stays in a drawer, the plan has no criterion for deciding, and every emergency is handled through improvised priorities. If a process does not appear in any decision, it has probably not really been considered.
A frequent mistake is thinking the BIA is a one-off exercise. It is instead a living document, to be compared with every test of the plan: if a test shows that a process takes longer than declared, the BIA must be corrected, and with it the linked strategy. An updated BIA is worth more than a perfect but stale one. The mistakes to avoid are these:
- Treating all processes as equally critical
- Doing the BIA only with IT, without operational managers
- Using precise, unverifiable figures instead of stated reasoning
- Not updating it after organisational or technological changes
- Forgetting suppliers and legal deadlines
8. The cluster's deep dives
This guide is dedicated to the BIA, but it sits within a group of continuity deep dives. The BIA answers the question of which processes are priorities. The other guides answer other questions: how to write the overall plan, how to structure the IT disaster recovery document, which recovery site model to choose, how to switch to a standby system and back, how to test the plan and how to map critical dependencies. Together they form the full preparation cycle, from diagnosis to verification.
To read in order of usefulness, start with the guide on the business continuity plan, which explains how the BIA fits into the rest of the document. Then read the mapping of critical dependencies, because it is the stage that most often changes the conclusions of the BIA. Only after that go into the detail of disaster recovery and testing, once you already have approved priorities and times to use as a reference.
9. How to start
This week, take a sheet and write the list of processes for a single function, the one that worries you most. For each process note the owner, the fixed deadlines and the systems it uses. Do not stop to assess the impact: the aim is to obtain a first verifiable list, which will then be discussed with the right people. Also write down the date on which you completed the sheet, because it will be needed to compare future versions.
Then ask the manager of that function to review it with you in a short meeting, and flag the activities that they cannot assign to anyone. Finally, ask your IT supplier which systems it knows for those processes and which data it has not yet catalogued. Also note the name of the person who answered, so you know whom to ask for confirmation later. With these three sources you will have the foundation of the first BIA.
Frequently Asked Questions
What is the difference between a BIA and a risk assessment?
A risk assessment looks for threats and the likelihood that they occur, whereas the BIA starts from the assumption that an event happens and measures what it means for the business. The two analyses complement each other: the assessment says what can happen, the BIA says how much each process matters when it does. Neither replaces the other, and in many companies they are carried out by the same people using different tools.
Who must approve the result of the BIA?
Management must approve the ranking of processes and the recovery objectives, because these are business decisions with costs and trade-offs. The IT manager validates technical feasibility, and the operational managers confirm the starting data. Keep the approval in writing, with date and version, because it will be the reference when you have to justify investments or compromises.
How long does a BIA take in a mid-sized company?
It depends on the number of processes, the availability of managers and the level of documentation already in place. A first version on a limited scope can be completed in a short working cycle, whereas an analysis extended to the whole organisation takes longer. It is best to start with the most critical functions and complete the rest in a second phase, rather than postponing everything.
Do I need dedicated software to carry out a BIA?
It is not essential. A well-structured spreadsheet is enough for most mid-sized companies, provided it includes fields for processes, owners, systems, dependencies, impacts and approved objectives. A dedicated tool can help keep history and versions, but it does not replace the interviews with managers, which are the most valuable part of the work.
How is the BIA turned into investment decisions?
The BIA shows which processes have the tightest recovery objectives. For those, you assess whether manual procedures are enough, or whether more frequent backups, data replication or a recovery site are needed. Compare the cost of each solution with the estimated impact of the interruption. There is no universal formula: the choice should be discussed with management and documented, stating the assumptions it rests on.
How often should the BIA be updated?
It should be reviewed at least once a year and whenever the organisation, systems or critical suppliers change significantly. It should also be updated after a real incident or a test exercise, if times emerge that differ from those declared. A quick check with the managers and connected systems can be done more often, without redoing the entire analysis.
Are qualitative estimates acceptable?
Yes, provided they are justified and approved. A qualitative rating such as low, medium or high, with a written explanation, is more useful than a precise figure that nobody can justify. When an auditor or a customer asks how you reached a conclusion, the written justification will be your answer.