In this lesson
After this lesson you can:
- Say what changes and what does not when a small organisation certifies.
- Solve the three structural problems that company size creates, rather than working harder at them.
- Decide a scope that keeps three years of cost and maintenance proportionate.
Read first: Whether you need ISO 27001 and how audit days are calculated
There is no size threshold.
The first thing to settle, because a lot of small companies assume otherwise and either delay or give up. ISO 27001 contains no minimum headcount, no revenue floor, and no small-organisation variant. A company of ten people is as eligible as a company of ten thousand.
It also contains no reduced version of itself. Clauses 4 to 10 apply in full. You still need a determined scope, a risk assessment with a method, a Statement of Applicability, documented information, competence and awareness arrangements, an internal audit, a management review and corrective action. All 93 Annex A controls still have to be considered, even if fewer end up applicable.
What does scale with size is the audit. Duration starts from the number of people doing work under your organisation’s control inside the scope, so a small team is at the bottom of the scale by construction, and there is correspondingly less evidence for an auditor to sample. How audit days are calculated covers the method, including why contractors count in that headcount.
What genuinely gets easier, and what does not.
| Aspect | Effect of being small | Why |
|---|---|---|
| Number of clauses you must satisfy | No change | Clauses 4 to 10 apply in full at any size. |
| Annex A controls to consider | No change | All are considered. Fewer may end up applicable. |
| Audit duration | Lower | Duration starts from headcount in scope. |
| Volume of evidence to produce | Lower | Fewer people, systems and changes to sample. |
| Getting everyone trained and aware | Easier | One session reaches the whole company. |
| Internal audit independence | Harder | Fewer people who did not build the thing. |
| Segregation of duties | Harder | One person often holds both ends of a process. |
| Continuity if the ISMS owner leaves | Harder | Often one person holds all the context. |
There are real advantages in the top half of that table, and they are worth claiming. Awareness training that reaches everyone in one session is genuinely easier than a programme for two thousand people. A risk assessment covering a handful of systems is tractable. You can change a process on Monday and have it operating by Tuesday, which is a speed no large organisation has.
The bottom half is where the work is, and none of it yields to effort. These are structural constraints created by having few people, which is why they need designing around rather than pushing through.
The three problems size actually creates.
Everything else about certifying a small company is ordinary work. These three are the ones that will come up at Stage 2 whatever else you do well, so solve them early and deliberately.
Problem one: internal audit independence.
Clause 9.2 requires you to audit your own ISMS at planned intervals, with auditors selected so the process stays objective and impartial. The rule underneath that is blunt: nobody audits their own work.
In a company of fifteen, the person who built the ISMS is usually the only person who understands it, which makes them simultaneously the obvious auditor and an impermissible one. Certification bodies see this constantly and look for it directly, because an audit signed by the implementer invalidates its own results however thorough the work papers are.
Two answers work at this size:
- Cross-functional auditing. Train people from different functions to audit across team lines. An engineer audits the HR security processes, an operations person audits engineering’s. Document independence per assignment, because auditors check the assignment matrix against the organisation chart.
- Buy internal audit as a service. External in delivery, internal in function: it remains your first-party audit. For many small teams this is cheaper than reorganising to create independence, and it has the side benefit of a competent first pass.
What does not work is an internal audit run two days before the certification visit, covering everything and finding nothing. That artefact tells an auditor the audit was performed for the certificate rather than for the organisation, and it invites deeper sampling everywhere.
Problem two: segregation of duties.
A.5.3 expects conflicting duties and areas of responsibility to be segregated. In a small team one person frequently deploys the code, grants the access, and reviews the logs, which is three conflicting duties in one pair of hands.
Two instincts here are both wrong. The first is to hire, which is a disproportionate response to a control. The second is to mark the control not applicable, which an auditor reads as evidence the control was not understood, and which then casts doubt on every other exclusion in your Statement of Applicability.
The correct route is compensating controls, and the control itself anticipates it. Document the conflict explicitly, then implement something that achieves equivalent assurance:
- Review after the fact by someone else, including a founder or director, with a record of the review.
- Logging the person performing the activity cannot alter or delete.
- Approval from outside the function for the highest-risk actions, even if the approver is not technical.
- Periodic independent checks, which can be part of the internal audit programme.
- Alerting on the specific actions that would be most damaging if misused.
Written down, that is a defensible answer. An auditor is testing whether you recognised the conflict and did something proportionate about it, not whether you have a headcount you do not need.
Problem three: a management review of four people.
Clause 9.3 specifies the inputs a management review must consider and requires documented information as evidence. It does not specify how many people attend or how long it takes.
So a small company’s review can legitimately be a short meeting of the founders and the ISMS owner. What it cannot be is the thing small companies actually do, which is to discuss security informally and often, decide things in passing, and have nothing recorded. That produces a genuine paradox at audit: a company that thinks about security constantly and can evidence having reviewed it never.
The fix is mechanical. Take the list of inputs from Clause 9.3, make it your agenda, work through it, and keep minutes that show each input was addressed and what was decided. An hour twice a year with real minutes beats continuous informal attention with none, and it is the single cheapest finding to avoid on this page.
Scope is your main lever.
If you take one decision seriously, take this one. Clause 4.3 lets you determine the boundaries of your ISMS, and that decision governs audit duration, evidence volume, and how much you maintain for three years.
Small companies get this wrong in a specific direction. They include everything, because a broad certificate feels more impressive and a narrow one feels like an admission. The result is three years of auditing and maintaining things nobody ever asked about, with two surveillance visits revisiting all of it before recertification.
Decide scope from the question your buyers are actually asking:
- What service or product are customers asking about when they raise security?
- Which systems hold or process their data?
- Which people touch those systems?
- Which legal entity contracts with those customers?
That is your scope. Write it as it will appear on the certificate, and widen it later when there is a reason. A narrow scope honestly stated is a better position than a broad one you cannot maintain, and presenting a certificate accurately is easier when the scope matches what you actually do.
What small teams over-build.
Effort spent here is effort not spent on the three real problems.
- Policy volume. Fourteen separate policies for fifteen people, mostly unread. A smaller set that people actually follow evidences better, because conformity is tested against your own documents and a policy you do not follow is a finding you wrote yourself.
- Enterprise risk methodology. A quantitative model with weighted factors for an organisation with twelve systems. Choose the simplest method you can apply consistently and defend.
- Formal committees. A security steering group with terms of reference that meets because the document says it does. The standard wants decisions made and recorded, not governance theatre.
- Tooling bought before the system exists. A platform is good at collecting evidence from a system that is running. It does not decide your scope or justify your exclusions.
What small teams under-build.
And the mirror image, which is where the findings actually come from.
- Records of routine activity. Small teams do the right things and record none of them. The access review happened in someone’s head. Stage 2 samples records, so the record is the control as far as the audit is concerned.
- Asset inventory. Especially cloud resources created by a script and SaaS tools adopted on a company card. An inventory listing only laptops is a reliable finding.
- Supplier management. A small company often depends on more third parties than a large one, proportionally, and has no register of them or of what each one holds.
- Leaver process evidence. You know the person left and you disabled the accounts. Showing both halves connected, for the same person on the same date, is a different thing.
- Key-person continuity. The biggest unmanaged risk in a small certified company. One person holds the whole ISMS in their head, and the first surveillance audit after they leave is a bad day. Record why decisions were made, not only what was decided, and involve a second person in the audit and review cycle.
The pattern across both lists is the same. Small organisations tend to build the documents and skip the records, when the audit weights it the other way round.
Want a scope that stays proportionate?
Tell us your headcount, what your buyers ask about, and who could audit whom. We will tell you the narrowest defensible scope, how to solve your independence and segregation problems at your size, and what audit duration to expect.
Key takeaways.
- There is no size threshold and no reduced version of the standard. Clauses 4 to 10 apply in full and all Annex A controls must be considered.
- Audit duration starts from headcount in scope, so small teams sit at the bottom of the scale with much less evidence to produce.
- Internal audit independence is the first structural problem. Use cross-functional auditing or buy internal audit as a service, and never let the implementer audit their own work.
- Segregation of duties is the second. Document the conflict and implement compensating controls. Do not hire, and do not mark A.5.3 not applicable.
- The management review is the third, and the cheapest to fix. Run Clause 9.3 as the agenda, keep real minutes, twice a year is fine.
- Scope is the lever that decides three years of cost. Decide it from what buyers ask about, write it as it will appear on the certificate, and widen it later.
- Small teams over-build policies, methodology and committees, and under-build records, inventory and supplier management. The audit weights records heavily.
- Key-person dependency is the biggest unmanaged risk once certified. Record why, not just what, and involve a second person in the cycle.
Quick check.
How does being a ten-person company change the ISO 27001 requirements?
- Fewer clauses apply
- Annex A is reduced to the essential controls
- The requirements are unchanged, and audit duration is lower
- Internal audit becomes optional
Check the answer
C. Clauses 4 to 10 apply in full at any size and all Annex A controls must be considered. What scales with headcount is audit duration and the volume of evidence available to sample. ISO/IEC 27006-1:2024
Your only DevOps engineer deploys code, grants production access and reviews the logs. What does A.5.3 expect?
- Hire a second engineer
- Mark the control not applicable
- Document the conflict and implement compensating controls
- Nothing, segregation applies only to larger organisations
Check the answer
C. The control expects conflicting duties to be addressed. Where separation is impractical, you record that and put in place review, immutable logging or outside approval that achieves equivalent assurance. Excluding the control because it is inconvenient is not a justification. A.5.3, segregation of duties
Spot the mistake. A fifteen-person company puts every entity, office and product in scope so the certificate looks comprehensive.
Check the answer
It bought three years of cost and maintenance it did not need.. Audit duration and maintenance effort both follow scope, and surveillance will revisit all of it twice before recertification. Scope should be decided from what buyers actually ask about, and widened later when there is a reason. Clause 4.3, determining the scope of the ISMS
Sources.
- ISO/IEC 27006-1:2024, requirements for bodies auditing and certifying an ISMSISO catalogue entry 82908, published March 2024Used here: That audit duration starts from the number of people doing work under the organisation’s control, which is why small teams sit at the bottom of the scale
- ISO/IEC 27001:2022, Clause 9.2, internal auditOur clause reference page, which carries the requirement and the impartiality constraintUsed here: The independence problem, and the two recognised answers for a small organisation
- ISO/IEC 27001:2022, Clause 9.3, management reviewOur clause reference page, which carries the mandated inputsUsed here: That a small review is acceptable and an unstructured or undocumented one is not
- ISO/IEC 27001:2022, control A.5.3, segregation of dutiesOur control reference page, which carries the control and its guidanceUsed here: That conflicting duties must be addressed, and that compensating controls are the route where separation is impractical
- ISO/IEC 17021-1, requirements for bodies auditing and certifying management systemsISO catalogue entry 61651Used here: That audit evidence comes from sampling, which is why fewer people means less to sample rather than fewer duties
Changes since publication
This page states no audit-day figures and no prices. Audit duration is governed by a table in Annex C of ISO/IEC 27006-1:2024, which is sold rather than published freely, so the mechanism is explained on our audit-days page instead and your certification body can give you your own number.
Last checked against ISO’s catalogue entries for the certification-body standards, and our own clause and control reference pages on 9 November 2026.