# Your security maturity survey

> Measure your security maturity in fourteen questions: a level per domain, an overall level, three priorities and the engagement level that fits.

Fourteen questions, one per domain, to place your security on five levels and know where to start. The result is computed in your browser: nothing is sent unless you decide to.

[Start the survey](https://www.cyviso.com/en/maturity-survey/#survey)

## Where your security stands

For each domain, choose the sentence that describes your situation today, not the one you aim for. A measure you could not show an auditor does not count.

On the web page, the result is computed in your browser from your answers. Nothing is sent.

### Which requirement weighs on you most today?

- The security requirements of our clients or our insurer
- NIS2, ISO 27001 certification or audits by large clients
- DORA and CSSF circulars

### Governance and compliance

- Security policies are missing, and nobody has listed the obligations that apply to us.
- A security policy is in force, and our regulatory and contractual obligations are listed.
- Indicators track security, audit logs are kept and regulatory returns are filed on time.
- Management steers risk with these indicators, and we anticipate what authorities and clients will expect.
- Compliance is checked continuously, and gaps surface on their own.

### Access and identity

- Accounts are shared, and access is not always removed when someone leaves.
- Multi-factor authentication protects at least the admin accounts, and someone approves each access.
- The main applications use single sign-on, and joiners and leavers open and close access automatically.
- Privileged access is managed separately, and access depends on context: device, location, risk.
- Rights are adjusted continuously, without manual steps.

### Continuity and backups

- Backups have never been restored to check they work, and there is no recovery plan.
- A backup policy exists, and a restore has been tested.
- A disaster recovery plan exists and has been tested.
- The continuity plan follows business priorities, and management has approved it.
- The infrastructure is redundant across sites, and recovery is rehearsed regularly.

### Incident readiness

- There is no incident response plan, and nobody to call in an attack.
- An incident response policy exists, and a response provider is under contract.
- Playbooks cover the likely cases, and tabletop exercises have tested them.
- A compromised device can be isolated automatically, and the traces an investigation needs are kept.
- Incident response is part of the company’s crisis management.

### Vulnerabilities and patching

- Vulnerability scans are occasional, with no deadline for fixes.
- Scans are scheduled, and fix deadlines are tracked.
- Scanning is continuous, and the great majority of patches is applied on time.
- Fixes are prioritised by real risk, and a dashboard follows them.
- Routine fixes apply automatically, and the remaining risk is measured.

### Workstations and devices

- Some devices are neither encrypted nor managed by the company.
- Devices are managed remotely and encrypted.
- Devices follow a hardened configuration, and users are not administrators.
- A non-compliant device is refused access to applications.
- No device is trusted by default: every access is checked.

### Detection and response

- Logs are read by hand, and basic antivirus is the only protection.
- An endpoint detection and response (EDR) tool covers part of the devices.
- EDR covers every device, and alerts are correlated.
- A SIEM centralises events, and threats are actively hunted.
- A security operations centre monitors around the clock, guided by threat intelligence.

### Vendors and third parties

- We have no vendor inventory and no security clauses in contracts.
- A vendor list exists.
- Critical vendors are identified, and their contracts include security obligations.
- Critical vendors are assessed, and our audit rights are exercised.
- Vendor risk is monitored continuously.

### Security awareness

- There is no structured training or awareness.
- New joiners receive security awareness.
- Phishing simulations run regularly.
- Training is tailored to roles: management, finance, development.
- Behaviour is measured, and awareness adjusts to the results.

### Cloud and infrastructure

- Default configurations are kept, with no overall view.
- A configuration baseline is the reference, checked by hand.
- Cloud configurations and infrastructure as code are scanned automatically.
- Policies written as code block deviations and fix them automatically.
- Infrastructure compliance is checked continuously.

### Secrets and API access

- Access keys are shared or written into code, with no vault.
- Secrets are kept in a vault, and rotation is manual.
- APIs are inventoried and protected by OAuth 2.0 and a web application firewall.
- An API gateway enforces the rules, and keys rotate under a policy.
- Connections are mutually authenticated (mTLS), and keys rotate automatically.

### Application security

- Security testing is occasional, and penetration tests are rare or absent.
- Code is scanned before each release, and an annual penetration test covers a limited scope.
- Code and dependency scanning is part of the build pipeline, and penetration tests follow the development cycle.
- Running applications are tested too, and a bug bounty programme sets fix deadlines.
- Testing is continuous, and routine fixes are automated.

### Use of AI

- We do not know which AI tools are in use, and no rule governs their use.
- The AI tools in use are inventoried, and an acceptable use policy is in force.
- Tools are classified by the data they may receive, and where data is processed is confirmed.
- AI risk is assessed when onboarding each vendor and in development.
- AI risk management is embedded and automated.

**How to read the result**

Each domain is placed on [five levels](https://www.cyviso.com/en/maturity-survey/#levels). “I don’t know” counts as level 1: what cannot be checked does not protect.

The overall level is the mean of the domains, rounded down.

**Privacy**

Your answers stay in your browser. Nothing is sent, unless you add the survey to a scoping call request.

## The five levels of the survey

Each level builds on the one before. Level 3 is where you can show evidence to an auditor or a client.

- **Ad hoc**: Security rests on people and reactions. What is done is neither written down nor checked.
- **Basic hygiene**: The essential measures are in place and documented, applied by hand.
- **Managed**: Measures are systematic, tooled and tested. You can show the evidence.
- **Proactive**: Risks are anticipated and steered with indicators. Gaps are closed before an incident.
- **Embedded**: Security is part of how the company runs, checked and adjusted continuously.

## Questions about the survey

### How is the survey result calculated?

Each answer places the domain on a level from 1 to 5, and “I don’t know” counts as level 1. The overall level is the mean of the thirteen domains, rounded down. The priorities are the three lowest domains.

### Are my answers sent to Cyviso?

No. The calculation happens in your browser. If you add the survey to your scoping call request, it goes into the message of the form, which you read before sending.

### Does the survey replace an audit?

No. The survey is self-declared, while an audit checks measures against evidence. It helps prepare the scoping call and set the first priorities.

### Which maturity level should we aim for?

It depends on your requirements. Level 3, Managed, lets you show evidence in the domains that clients, insurers and auditors examine. Entities subject to DORA have explicit requirements on continuity, incidents and ICT providers.

## Let’s talk about your survey

Add your survey to the request, or simply describe your situation. The scoping call sets the priorities and the engagement level that fits.

[Book a scoping call](https://www.cyviso.com/en/maturity-survey/#contact)
