
SABSA Framework: What It Is, Its 6 Layers, and How to Use It
The SABSA framework is a business-driven method for designing enterprise security architecture. It starts with what the business needs to protect and why, then traces every security control back to a business goal, so security decisions can be explained in business terms rather than only in technical terms.
SABSA stands for Sherwood Applied Business Security Architecture. Developed in the 1990s by John Sherwood and first published in 1995, it remains one of the most widely used security architecture frameworks in banking, government and healthcare. Unlike a checklist of controls, SABSA is a way of thinking: security exists to enable the business safely, not to block it.
What Is the SABSA Framework?
At its core, SABSA answers four questions for any organisation: what assets matter, what risks threaten them, what security services are needed, and how those services will be delivered and measured. It treats security architecture as a layered design problem. Each layer asks the same six questions — what, why, how, who, where and when — but at a different level of detail, from the boardroom to the wiring closet.
That traceability is the framework’s defining strength. If a chief executive asks why the company spends money on a particular control, a SABSA-based architecture can show the chain: business objective, business risk, security policy, security service, mechanism and finally the product that delivers it. Nothing is justified by habit or fashion alone.
SABSA is also deliberately independent of technology and vendors. It does not tell you which firewall or identity product to buy. It tells you how to decide, document and defend the choice, and how to revisit it when the business or the threat landscape changes.
The Six Layers of SABSA
SABSA is usually described as six layers, moving from the most abstract to the most concrete. Each layer is owned by a different audience, which is why the framework works well for communicating across an organisation.
| Layer | Focus | Main audience | Example output |
|---|---|---|---|
| 1. Contextual | Business context and goals | Board and executives | Business attributes and risk appetite |
| 2. Conceptual | Business security concepts | Business managers | Security policies and domain concepts |
| 3. Logical | Security services and relationships | Security architects | Logical security architecture |
| 4. Physical | Security mechanisms | Engineers and designers | Mechanism designs and standards |
| 5. Component | Products and tools | Implementers and vendors | Product selections and configurations |
| 6. Operational | Running and measuring security | Operations teams | Procedures, monitoring and metrics |
1. Contextual layer
This is the business view. Architects work with executives to identify business drivers, critical assets, stakeholders and the organisation’s appetite for risk. A hospital’s contextual layer, for example, would emphasise patient safety, confidentiality of medical records and uninterrupted care. A bank’s would emphasise trust, regulatory compliance and uninterrupted payments.
2. Conceptual layer
Here the business needs become security concepts: the policies, risk models and security domains that will govern the organisation. A domain is a group of elements that share the same security policy — for example, a card-payment domain or a research-and-development domain. The conceptual layer also introduces trust: who is trusted, for what, and under which conditions.
3. Logical layer
The logical layer defines the security services required, such as identification and authentication, authorisation, audit, confidentiality and integrity, without yet choosing how they will be built. It maps the relationships between services, users and information, and it is where architects decide what must be separated, what must be monitored and what must be recoverable.
4. Physical layer
The physical layer turns services into mechanisms: a single sign-on mechanism, a network segmentation mechanism, an encryption mechanism. It specifies how each service will actually work, the standards it must meet and how mechanisms interact. Technology choices begin here, but products are still not named.
5. Component layer
Only at this layer are specific products, tools and configurations selected — a particular identity provider, a named firewall model, a specific encryption library. Because the earlier layers justify the choice, procurement becomes a traceable decision rather than a sales-driven one, and replacing a product later does not require redesigning the architecture.
6. Operational layer
The operational layer covers the day-to-day running of security: procedures, staffing, monitoring, incident response and measurement. It asks how the organisation will know the architecture still works, and what happens when it does not. Feedback from operations feeds back into the higher layers, so the architecture improves rather than fossilising.
The SABSA Matrix Explained
The six layers form the rows of the SABSA matrix. Its columns are the six questions adapted from the Zachman framework: what (assets), why (motivation), how (process), who (people), where (location) and when (time). Every cell of the matrix is a viewpoint that must be addressed.
For example, the “what” column at the contextual layer lists business assets; at the logical layer it lists information assets and their classifications; at the component layer it lists the actual data stores and devices. Reading down a column shows how one concern is refined at every level of detail. Reading across a row shows a complete view at one level of abstraction.
In practice, few teams fill in all 36 cells on the first pass. The matrix is most useful as a completeness check: if a column is empty at a layer, something has probably been forgotten. A security design that specifies technology (how) without identifying who may use it and where is incomplete, and the matrix makes that gap visible.
SABSA vs Other Frameworks
SABSA is often compared with other architecture and security frameworks. It is not a replacement for them; in many organisations it sits above them, giving business meaning to the controls they describe.
| Framework | Main purpose | How it relates to SABSA |
|---|---|---|
| TOGAF | Enterprise architecture generally | TOGAF describes the whole enterprise; SABSA adds the security architecture view within it |
| Zachman Framework | Classifying architecture artefacts | SABSA’s matrix is inspired by Zachman, but adds business risk and security services |
| MITRE ATT&CK | Catalogue of attacker tactics and techniques | Use ATT&CK to test whether SABSA’s controls cover the techniques attackers actually use |
| ISO 27001 | Information security management system | SABSA can design the architecture that ISO 27001 controls then govern and audit |
| NIS2 Directive | EU cybersecurity obligations | SABSA traceability helps show regulators why each NIS2 measure exists and who owns it |
How to Implement SABSA: A Worked Example
Consider a regional bank that wants to launch online account opening. Its business objective is to win new customers without a branch visit. Its risks include identity fraud, account takeover and regulatory fines.
- Capture the business context. Interview executives and product owners. Record the objective, the assets (customer identity data, account records), the stakeholders and the risk appetite — for example, zero tolerance for unauthorised account access.
- Define the conceptual layer. Write the policy: every new account must be verified to a defined assurance level, and no account may be opened without an audit trail. Define the security domain for customer onboarding and its trust rules.
- Design the logical services. Specify identity proofing, authentication, authorisation, audit and fraud-monitoring services, and how they relate to each other and to the core banking system.
- Choose the mechanisms. Decide on document verification, liveness checks, one-time passcodes, step-up authentication for high-risk cases and immutable audit logs.
- Select the components. Only now evaluate vendors against the mechanism requirements, cost and integration effort. Record the decision and its rationale.
- Plan operations. Define who monitors onboarding fraud, how quickly suspicious accounts are frozen, which metrics are reported to management, and how the service is tested.
- Review and trace back. Walk the chain from a component — say, the liveness-check product — back up to the business objective. If a link is missing or weak, fix the architecture before launch, not after an incident.
This worked example adds what many short explainers miss: SABSA is not a document, it is a chain of reasoning that can be audited. A team that can walk that chain in either direction — from business goal down to product, or from product up to business goal — has a genuine security architecture. A team that cannot is holding a pile of products.
Benefits and Limitations
The main benefits are better board-level communication, clearer prioritisation, easier compliance evidence and more resilient design. Because controls are tied to business risk, security spending can be defended and, just as importantly, spending that does not reduce a business risk can be challenged. The framework also scales: a small organisation can use a light version of the matrix, while a large one can formalise every cell. The SABSA Institute offers Foundation and Practitioner certifications for architects who want formal recognition, but certification is not required to use the framework.
Frequently Asked Questions
What does SABSA stand for?
SABSA stands for Sherwood Applied Business Security Architecture, named after its originator John Sherwood. It is a framework for building security architecture from business requirements rather than from technology.
What are the six layers of SABSA?
The six layers are contextual, conceptual, logical, physical, component and operational. They move from business goals to the products and procedures that deliver security, with each layer refining the one above it.
Is SABSA a security standard?
No. SABSA is a methodology and architecture framework, not a certifiable standard like ISO 27001. It helps organisations design their security architecture; standards and regulations then provide the requirements and audit criteria that architecture must satisfy.
How is SABSA different from TOGAF?
TOGAF is a general enterprise architecture framework covering business, data, application and technology architecture. SABSA focuses specifically on security architecture and can be used inside a TOGAF programme to provide the security view.
Is SABSA still relevant in 2026?
Yes. As organisations adopt cloud, artificial intelligence and zero-trust designs, the need to trace security choices back to business risk has grown, not shrunk. SABSA’s layering and traceability apply just as well to cloud-native services as to on-premises systems, even if some of its language feels dated.
Conclusion
The SABSA framework turns security from a technical shopping list into a business argument. By working through its six layers and its matrix of questions, an organisation can explain every control, prioritise what matters, and adapt as the business changes. Start small, pick one important business service, trace one chain from goal to control, and let the value of that traceability make the case for going further. For the access decisions that sit inside any such architecture, read our guide to role-based access control (RBAC) next.


