Risk and vulnerability analysis in 7 steps, with template

Back to Blog

Last updated: 23 October 2026 · By Gaute Wangen

A risk and vulnerability analysis (in Norwegian risiko- og sårbarhetsanalyse, or ROS-analyse) gives you an overview of what can go wrong, how likely it is, what it would cost, and which measures have the most effect. Here we go through the method in seven steps, with a concrete example and a free template you can use straight away.

What is a risk and vulnerability analysis?

A risk and vulnerability analysis is a systematic review of unwanted events that could affect an organisation, a system or a process. For each event, you assess the likelihood that it happens and the impact if it does. Together they give a risk level, and the risk level decides which measures to prioritise.

The vulnerability part is about how well you cope with an event once it happens: how quickly you detect it, how long you can manage without a system, and how well your preparedness works.

In Norway, where it is called ROS-analyse, the method is particularly common in the public sector and in municipalities. The method builds on the same principles as risk assessment under ISO 31000, ISO/IEC 27005 and the Norwegian standard NS 5814.

When do you need a risk and vulnerability analysis?

Many organisations are required to assess risk, even if the rules do not always use the term risk and vulnerability analysis:

  • Municipalities must have an overall risk and vulnerability analysis under the Civil Protection Act (sivilbeskyttelsesloven), and risk and vulnerability analyses are required in land-use planning under the Planning and Building Act (plan- og bygningsloven).

  • Organisations covered by the Digital Security Act must assess the risk to their network and information systems and use the assessment to choose security measures. Read more in Norway's Digital Security Act: 7 requirements to meet.

  • Anyone processing personal data must have a level of security appropriate to the risk (GDPR Article 32), and must carry out a data protection impact assessment (DPIA) when the processing is likely to result in a high risk.

  • Organisations that are certified, or want to be certified, to ISO/IEC 27001 must have a documented process for risk assessment and risk treatment.

See also: Records of processing activities: guide and free template.

A risk and vulnerability analysis also makes sense before you introduce a new system, change supplier, move services to the cloud or make major organisational changes.

Risk and vulnerability analysis, risk assessment and risk analysis: what is the difference?

The terms are often used interchangeably. Under NS 5814 and ISO 31000, they relate like this:

  • Risk assessment is the whole process: identifying, analysing and evaluating risk.

  • Risk analysis is the part where you find causes and assess likelihood and impact.

  • Risk evaluation is the part where you compare the risk with the acceptance criteria and decide what needs treatment.

  • Risk and vulnerability analysis (ROS-analyse) is a risk assessment that also looks at vulnerability and preparedness. It is the established approach in Norway.

So the method is the same. What matters most is that you are consistent, and that everyone involved understands the terms the same way.

How to do a risk and vulnerability analysis in seven steps

1. Define the scope and decide the criteria

Decide what to analyse and what to leave out. It could be the whole organisation, one service, one IT system or one supplier. An analysis that is too broad becomes superficial. One that is too narrow misses the dependencies.

Before you start, also decide:

  • Scales for likelihood and impact, for example 1–5 with concrete descriptions (see the template).

  • Acceptance criteria: the risk level management accepts without new measures.

  • Participants: people who know the service, the system and the users. An analysis done by one person at a desk is rarely good.

2. Describe what you are analysing

Write a short description of what you are analysing: which services and processes it supports, what information it handles, which IT systems and suppliers it depends on, and who uses it. Dependencies are often where the biggest risks hide. Read more in What else stops? Understanding IT system and supplier dependencies.

3. Identify unwanted events

List events that could threaten the service or the information. Make each event concrete, for example "An employee's email account is taken over after phishing", not just "Cyber attack".

Good sources are previous incidents and non-conformities, threat assessments from the Norwegian National Security Authority (NSM) and your industry, checklists, and structured brainstorming. A useful technique is to work through guide questions such as "What if the system is unavailable?", "What if the information goes astray?" and "What if the information is changed without us noticing?". That way you cover availability, confidentiality and integrity.

4. Find causes and existing barriers

For each event: what could cause it, and what do you already have in place that prevents it or limits the damage? Existing barriers must be part of the assessment. Otherwise you overestimate the risk and spend money on measures you already have.

For events with many causes and consequences, a bowtie analysis is a good tool. It shows the causes on one side, the consequences on the other, and the barriers in between.

5. Assess likelihood and impact

Give each event a value for likelihood and one for impact, based on the scales from step 1. Justify the assessment in a sentence or two. The justification is often worth more than the number, because it makes the assessment verifiable next time.

Place the events in a risk matrix. That quickly shows which risks stand out.

5 x 5 risk matrix with likelihood and impact, divided into low, medium, high and very high risk

The template uses likelihood × impact and four levels:

Risk score

Level

Typical follow-up

1–4

Low

Acceptable. Keep an eye on it.

5–9

Medium

Consider measures if the cost is reasonable.

10–15

High

Plan measures, with an owner and a deadline.

16–25

Very high

Not acceptable. Measures must be put in place, and management must be informed.

6. Evaluate the risk and choose measures

Compare each risk with the acceptance criteria. For risks above the limit, choose one or more strategies:

  • Reduce: put in measures that lower the likelihood or the impact.

  • Avoid: stop the activity or choose a different solution.

  • Transfer: share the risk through contracts or insurance.

  • Accept: accept the risk deliberately, with a justification that management stands behind.

Then assess the residual risk, the risk that remains after the measures are in place. A measure without an owner and a deadline is rarely carried out, so give every measure both.

7. Document, get management approval and follow up

Management owns the risk and must approve both the result and the residual risk. Follow up the measures regularly, and update the analysis when something changes: new systems, new suppliers, new threats or an incident that has actually happened. An analysis that sits in a folder for three years offers little protection.

Example: risk and vulnerability analysis of cloud email and file storage

The example is a medium-sized organisation that uses a cloud office suite for email, file storage and sharing. The scales are the same as in the template.

#

Unwanted event

Main causes

L

I

Risk

Measures

Residual risk

1

A user account is taken over

Phishing, reused passwords, some users without multi-factor authentication

4

4

16 Very high

Multi-factor for everyone, block legacy sign-in protocols, phishing training

8 Medium

2

Ransomware encrypts shared files

Infected attachment, sync spreads the encrypted files

3

5

15 High

Backup outside the cloud service, restore test twice a year

6 Medium

3

An email with personal data goes to the wrong recipient

Address autocomplete, time pressure

4

3

12 High

Warning when sending externally, routine for sensitive attachments, training

6 Medium

4

A former employee still has access

No fixed routine for closing accounts when people leave

3

3

9 Medium

Leaver routine linked to HR, quarterly access review

3 Low

5

Files are shared with links anyone can open

Default sharing setting, lack of awareness

3

3

9 Medium

Restrict external sharing, expiry dates on sharing links

4 Low

6

The service is unavailable for several hours

Outage at the supplier

2

3

6 Medium

Accepted. Contingency plan and contact list for outages

6 Medium

L = likelihood, I = impact.

Note event 6: the risk is accepted, but with a justification and a simple contingency plan. That is a valid decision, as long as it is made deliberately and documented.

Risk and vulnerability analysis template

Download the risk and vulnerability analysis template (Excel)

The template has four tabs:

  1. Guidance: the seven steps in short form.

  2. Scales: likelihood and impact from 1 to 5, with descriptions for finances, privacy, reputation and service delivery. Adapt the descriptions to your organisation before you start.

  3. Analysis form: one row per unwanted event, with causes, existing barriers, likelihood, impact, risk, measures, owner, deadline and residual risk. Risk and risk level are calculated automatically.

  4. Risk matrix: shows how many events fall in each cell, before and after measures.

Common mistakes in risk and vulnerability analyses

  • Events that are too broad. "Cyber attack" is not an event you can assess. "Ransomware encrypts the file server" is.

  • Existing measures are forgotten. The risk ends up too high and the list of measures too long.

  • Numbers without justification. Next time, nobody knows why the likelihood was set to 4.

  • Measures without an owner and a deadline.

  • The analysis is never updated. The risk picture changes faster than most people think.

  • Management is not involved. Accepting risk is a management decision, not one for the IT department.

Want to go a step further? The risk and vulnerability analysis gives you the overview. A bowtie analysis goes deep into one unwanted event and shows the causes, consequences and barriers around it. Read more in Bowtie Analysis: A Step-by-Step Guide with Example.

From spreadsheet to a living risk register

A spreadsheet works well for the first analysis. It gets heavy when you have many analyses, many measures to follow up, and a need to show management and the auditor how the risk develops over time.

In Cyrigo, you do the risk and vulnerability analysis in a structured process based on ISO/IEC 27005. You can start from a template, use a wizard, or import an existing analysis from Excel. Unwanted events can be analysed with bowtie, the risks are gathered in one shared risk register with a risk matrix, and the measures get an owner and a deadline and are followed up in the platform. The risk picture shows current and planned risk side by side, so management can see the effect of the measures. Cyrigo is developed in Norway and hosted in Scandinavia.

See how it works on the product page, or book a demo.

Risk and vulnerability analysis: frequently asked questions

What does ROS mean?

In Norway, a risk and vulnerability analysis is called ROS-analyse. ROS stands for risiko og sårbarhet, risk and vulnerability.

How long does a risk and vulnerability analysis take?

A focused analysis of one system or one service can be done in one or two workshops with the right people. An overall analysis of a whole organisation or municipality takes considerably longer.

How often should the analysis be updated?

At least once a year, and always after significant changes: new systems, new suppliers, reorganisation or a serious incident.

Who should take part?

The people who know the service and the systems: the service owner, users, IT, security and preferably the data protection officer. Someone who leads the process and knows the method makes the work more efficient.

Is a DPIA the same as a risk and vulnerability analysis?

No. A risk and vulnerability analysis looks at risk to the organisation. A DPIA looks at risk to the people the data is about. The methods are similar, and much of the work can be reused, but a DPIA has its own requirements under GDPR Article 35.

Back to Blog