Part of ISO 27001: Critical success factors, a series on building a management system that reflects how your organisation really runs.
What it is
Risk analysis is how you work out what could go wrong within your scope, how much it matters and who owns it. The risk register is where the answer is recorded.
Most managers love talking about risk; it's their day job. What they don't love is spending time on documents that don't feel like they add value. A good risk process starts with that conversation and writes it down well.
What the standard asks
Clause 6.1.2 asks you to define and apply a risk assessment process that:
- sets criteria for assessing and accepting risk
- produces consistent, valid and comparable results when repeated
- identifies risks to the confidentiality, integrity and availability of information in scope
- names a risk owner for each risk
- assesses the likely consequences and likelihood, and determines a level of risk
- compares results against your criteria and prioritises them for treatment
Clause 8.2 asks you to carry out the assessment at planned intervals and whenever significant changes happen, and to keep the results. The current standard doesn't require an asset-by-asset approach; you're free to assess risk in the way that suits your organisation.
What good looks like
- A short register. Ten to twenty well-worded risks is a good starting point for a first workshop. The final number follows your scope and exposures, not a target. A short register gets read; a long one gets maintained by one person.
- Risks written as outcomes the business would recognise, with a named owner who agrees it's theirs.
- An agreed method with impact tables that cover more than one kind of harm (financial, operational, regulatory, reputational) and acceptance criteria leadership have signed off.
- Scores treated as informed opinion. Where a risk sits on the heat map matters more than its number; a likelihood-1, impact-5 risk calls for a very different conversation from a likelihood-5, impact-1 one.
- A live register. It changes when events, metrics or the business tell you something new, not just once a year.
If your executives don't care about the risks in the register, the risks simply aren't connected to the real issues yet.
Evidence an auditor will ask for
- The documented risk assessment method, including acceptance criteria
- The risk register, with owners, scores and review dates
- Records of when assessments were carried out and what triggered them
- Minutes showing the forum reviewed the results and owners accepted residual risk
Common failure modes
- The template register. Two hundred generic risks downloaded from somewhere. Nobody owns them, nobody reads them, and the auditor can tell.
- Control gaps written as risks. "No MFA on email" is a finding; the risk is what happens to the business if accounts are taken over. Writing the outcome is what lets leadership weigh it.
- Security owns everything. Risk owners are the people who can actually make decisions about the risk, which is usually the business, not the security team.
- False precision. Elaborate scoring formulas that suggest more certainty than anyone has. Honest partial data is better than fully populated fiction.
How it connects
Feeds from
- The agreed scope and context.
- Security events, which show what actually happened.
- Metrics, which show whether controls are holding up.
Feeds into
- The management forum, which reviews and challenges the results, while each risk owner accepts the residual risk for their own risks.
- The risk improvement plan, which records what you'll do about each risk.
- Control selection, which starts from the risks you need to treat.
Where it sits in the plan
Stage: Plan. Risk assessment is at the heart of the first Plan. Agree the method, then sit down with managers, pen and paper, and ask what they're worried about and where they'd like more clarity. Come back with suggested wording. On later cycles, events and metrics drive targeted updates between the planned reviews.
The plan in the main article shows how this fits with everything else, and the bootstrapping guide covers each stage in detail.
Resources
Tom Whipp is the founder of RiskQuilt. He spent around thirty years in cyber and technology risk in UK financial services, including CISO and Director of Technology Risk roles.