Explore our new Cybersecurity Terms Glossary, explained by our experts.

View the Glossary View the Glossary
PRIVIA

Cybersecurity Terms Glossary

What Is BSIMM?

BSIMM is a descriptive model of software security built from observed practice in real firms. It measures what your programme does against what peers do, rather than against a list of requirements.

Illustration for the cybersecurity term BSIMM
Home / Blog / Cybersecurity Terms / What Is BSIMM?
17 November 2026

What Is BSIMM?

What Is BSIMM?

BSIMM, the Building Security In Maturity Model, is a measuring stick rather than a rulebook. It began in 2008 when Gary McGraw, Sammy Migues, and Brian Chess interviewed nine firms about what their software security teams actually did, wrote down the activities they observed, and published the result. Every edition since has repeated that exercise with a larger pool of participants, and the model changes when observed practice changes.

That is the distinction worth holding on to. Most frameworks are prescriptive: here is what you should do. BSIMM is descriptive: here is what firms like yours are doing, and here is where you sit among them.

How the Model Is Structured

Activities are organised into four domains, each containing three practices, giving twelve practices in total.

DomainPractices
GovernanceStrategy and metrics, compliance and policy, training
IntelligenceAttack models, security features and design, standards and requirements
SSDL touchpointsArchitecture analysis, code review, security testing
DeploymentPenetration testing, software environment, configuration and vulnerability management

Within each practice, individual activities are graded at three levels. Level 1 activities are common, level 3 activities are rare and usually appear only in mature programmes. An assessment records which activities an organisation performs, producing a scorecard that can be compared with the pool average and with a chosen vertical such as financial services, healthcare, or technology.

What an Assessment Involves

A BSIMM assessment is an interview exercise, not a technical audit. An assessor speaks to the software security group, engineering leadership, and development teams, then marks each observed activity. There is no pass mark and no certificate; the deliverable is a spider chart and a set of observations.

Two outputs tend to be useful. The first is the gap between what the governance function believes happens and what engineers describe, which is often wide. The second is the comparison with peers, which changes budget conversations: it is easier to request investment in architecture analysis when the data shows most comparable firms already do it.

BSIMM Compared With OWASP SAMM

The two are frequently confused. OWASP SAMM is prescriptive and open: it sets out five business functions with maturity levels you can work towards, and the materials are free to download and self-assess. BSIMM reflects what a measured community already does, with data collected by the model’s owners.

In practice, teams use SAMM to plan and BSIMM to benchmark. A security programme that needs a roadmap gains more from SAMM. One that needs to justify its position to an executive committee or a prospective client gains more from a BSIMM scorecard. Neither replaces hands-on verification, which is why both models include penetration testing and code review as distinct practices.

Where It Fits in a UK Programme

BSIMM carries no regulatory status in the UK, so nobody will ask for it in a tender the way they ask for Cyber Essentials or ISO 27001. Its value is internal, and it maps neatly onto obligations that do carry weight. The NCSC’s secure development and deployment guidance covers much of the same ground in principle form, and software supply chain expectations arriving through the EU Cyber Resilience Act will push UK manufacturers selling into Europe towards exactly the activities BSIMM records: threat modelling, vulnerability handling, and coordinated disclosure.

Watch for two failure modes. Treating the scorecard as a target invites activity theatre, where teams perform a level 3 activity badly to tick a box. And benchmarking against the pool average is reassurance, not a risk statement: the average firm in any pool still ships exploitable defects. Use the model to find the practices you have genuinely skipped, then fix those.

Getting Value Without a Formal Assessment

You do not need a commercial engagement to benefit. Take the twelve practices, walk them with your engineering leads, and mark honestly which activities happen on every release rather than occasionally. Most organisations discover the same three gaps: nobody owns attack modelling, architecture review happens only for new systems, and security testing stops at the scanner. Those are tractable problems with real returns.

If you want independent evidence of how your software actually stands up, our web application security services provide the testing side of the picture that a maturity model can only describe.

FAQ – Frequently Asked Questions

Is BSIMM a Certification?

No. BSIMM produces a descriptive scorecard showing which activities an organisation performs and how that compares with other participants. There is no pass mark, no certificate, and no accredited body issuing one. Organisations sometimes share assessment results with clients as evidence of software security maturity, but it carries no formal status in procurement the way ISO 27001 certification does.

What Is the Difference Between BSIMM and OWASP SAMM?

BSIMM is descriptive and tells you what observed firms actually do, so it suits benchmarking. OWASP SAMM is prescriptive and open source, setting out maturity levels you can work towards, so it suits planning and self-assessment at no cost. Many teams run a SAMM self-assessment to build a roadmap and commission BSIMM when they need external comparison.

How Long Does a BSIMM Assessment Take?

Typically a few weeks of elapsed time for a single business unit, built around interviews with the software security group, engineering managers, architects, and developers, plus documentation review. Large organisations assessing multiple divisions take longer. The scoring itself is quick; the useful work is reconciling what different parts of the organisation believe is happening on every release.

Do Small Development Teams Benefit From BSIMM?

The full model assumes a dedicated software security group, which a ten-person team will not have. The practice list is still useful as a checklist: ask which of the twelve practices has a named owner and happens consistently. Smaller teams usually get faster value from OWASP SAMM or from fixing a specific gap such as dependency management or threat modelling.

Related Terms

You May Be Interested In These

What Is Kerberos Authentication?

Cybersecurity Terms

17 November 2026

What Is Kerberos Authentication?

Read More
What Is RPKI?

Cybersecurity Terms

17 November 2026

What Is RPKI?

Read More
What Is the CISA Certification?

Cybersecurity Terms

17 November 2026

What Is the CISA Certification?

Read More