Bowtie Analysis: A Step-by-Step Guide with Example
Back to Blog
Last updated: 23 October 2026
Bowtie analysis is a risk assessment method that shows, in one diagram, how an unwanted event can happen, what it can lead to, and which barriers stop it or limit the damage. This guide explains the parts of a bowtie, walks through a bow-tie risk assessment step by step, and shows a complete example from information security.
What is bowtie analysis?
A bowtie diagram puts one unwanted event in the middle. The causes that can lead to the event are on the left, and the consequences it can lead to are on the right. Between them are the barriers: on the left, the ones that prevent the event, and on the right, the ones that reduce the consequences once it has happened. The finished diagram has the shape of a bow tie, which gives the method its name.

The strength of the method is that it shows how many different causes and consequences a single event can have, and which barriers protect against each of them. That makes it easier to see where you are well protected, where you depend on a single barrier, and where a new measure gives the most effect.
The method is usually traced to hazard analysis in the chemical industry in the late 1970s, and it became widespread in oil and gas in the 1990s. Today it is used in aviation, healthcare, finance and, increasingly, information security.
The parts of a bowtie diagram
Element | What it means | Example |
|---|---|---|
Hazard | Something with the potential to cause harm, which you need in order to operate | Customer data stored in the cloud |
Top event | The moment you lose control of the hazard | Unauthorised access to customer data |
Threats (causes) | What can lead to the top event | Phishing, weak passwords, misconfiguration |
Preventive barriers | Measures that stop a threat from causing the top event | Multi-factor authentication, access reviews |
Consequences | What the top event can lead to | Personal data breach, service disruption |
Mitigating barriers | Measures that reduce or stop the consequences | Logging and alerting, incident response, backups |
Escalation factors | Conditions that weaken a barrier | Access reviews skipped when the team is busy |
In information security, the hazard is usually an asset you depend on: information, a system or a service. The top event is a loss of confidentiality, integrity or availability for that asset.
How to do a bow-tie risk assessment, step by step
1. Choose the top event
Start with one specific event, worded as the moment control is lost. "Unauthorised access to customer data" works. "Cyber attack" is too broad, and "Phishing" is a cause, not an event. If you already have a risk assessment, pick the risks with the highest score or the most uncertainty.
2. Identify the threats
List everything that can lead to the top event. Use incident history, threat reports and the people who run the system. Keep each threat independent: each one should be able to cause the event on its own.
3. Identify the consequences
List what the event can lead to: for the organisation, its customers and the people whose data is involved. Think about cost, legal duties, operations and reputation.
4. Add the preventive barriers
For each threat, list the barriers that already exist and stop it from causing the event. A barrier should be specific and something you can check. "Security awareness" is vague. "Phishing training twice a year, with simulated phishing tests" is a barrier.
5. Add the mitigating barriers
For each consequence, list the barriers that limit the damage once the event has happened: detection, response, recovery and communication.
6. Add escalation factors and owners
Ask what could make each important barrier fail, and add those escalation factors to the diagram. Give every barrier an owner, so it is clear who keeps it working.
7. Assess the risk and decide on treatment
Estimate the likelihood of the top event with the current barriers, and the severity of the worst realistic consequence. Compare the result with your risk acceptance criteria. Where the risk is too high, look for threats with only one barrier, or no barrier at all, and consequences with weak mitigation. That is usually where a new measure pays off most.
Example: unauthorised access to customer data

A mid-sized company keeps its customer data in a cloud service. The bowtie gives this picture:
Threats: phishing that steals user credentials, weak or reused passwords, misconfigured cloud storage, and a former employee who still has access.
Preventive barriers: multi-factor authentication for all users, phishing training, a password manager, blocking legacy sign-in, configuration reviews, least privilege, a leaver process linked to HR, and quarterly access reviews.
Consequences: a personal data breach, service disruption, fines and financial loss, and loss of customer trust.
Mitigating barriers: logging and alerting with round-the-clock monitoring, an incident response plan, tested backups, data loss prevention, cyber insurance, breach notification to the data protection authority within 72 hours as GDPR requires, and a crisis communication plan.
Escalation factor: access reviews are skipped when the team is busy.
The diagram shows two things a risk list would not. First, the former-employee threat depends on processes that are easy to skip, so it is weaker than it looks on paper. Second, several consequences depend on the same mitigating barrier: whether anyone sees the alerts. Making monitoring reliable outside office hours protects against several consequences at once.
Bowtie analysis compared with other methods
Risk matrix or risk register: good for getting an overview of many risks. A bowtie goes deeper into one risk at a time. The two work well together: use the register to find the risks that matter, and the bowtie to understand and treat them.
Fault tree analysis: a detailed, often quantitative, analysis of the causes of one event. The left side of a bowtie is a simplified fault tree.
Event tree analysis: follows how an event develops and which barriers succeed or fail. The right side of a bowtie is a simplified event tree.
Bowtie works well for the 5–10 risks that matter most, where you need to explain the risk picture to management, auditors or people outside the security team.
Bowtie in the Cyrigo platform

In Cyrigo, bowtie diagrams are part of the risk assessment, which follows ISO/IEC 27005. There, a risk scenario is a combination of assets, vulnerabilities, threats, controls and consequences.
You can create several bowtie diagrams in one assessment, and reuse the same treatment across causes and consequences. You can work on one cause or consequence at a time, extract individual risks from the diagram, and compare the cost of a treatment with the risk reduction it gives.

You can start an assessment from an empty template, from a template for the scope you are assessing, or with a wizard. Assessments you already have in Excel or in reports can be brought in with the row importer.
Risks you extract from a bowtie go into the risk matrix and a shared risk register. There you can see inherent and residual risk, and follow up treatments with owners and deadlines.

The risk picture compares current and planned risk, and shows which treatments are recommended. That makes it easy to show management what the planned measures will achieve.

Cyrigo is built in Norway and hosted in Scandinavia. See the Product page for the rest of the platform, or book a demo.
Frequently asked questions about bowtie analysis
What is the difference between bowtie and bow-tie?
None. Both spellings, and "bow tie", refer to the same method.
Is bowtie analysis qualitative or quantitative?
Usually qualitative, but it can be combined with likelihood and consequence scores, or with quantitative fault tree and event tree analysis where you have the data.
How many threats and consequences should a bowtie have?
Most useful diagrams have between three and eight of each. If you have many more, the top event is probably too broad and should be split.
Can bowtie be used in information security?
Yes. It fits well with ISO/IEC 27005 and with the risk assessments required by ISO/IEC 27001, NIS2 and GDPR, because it shows clearly which controls protect against which threats.
Related reading: What is GRC? Governance, risk and compliance explained · What else stops? Understanding IT system and supplier dependencies
Back to Blog