“Access management is inadequate.” This sentence can’t be fixed or disputed. It puts the auditee on the defensive, leaves the corrective action vague and turns the closing meeting into an argument about evidence. The value of an audit is measured less by the evidence collected in the field than by the quality of the finding written in the report. Below, for internal auditors and ISMS owners, we explain the criterion–condition–cause–effect–recommendation structure, common wording mistakes with their corrected versions, the rating scale and agreeing the finding with the auditee.
Elements of a good finding
ISO 19011:2018 defines an audit finding as the result of evaluating the collected audit evidence against audit criteria. The definition already gives the two mandatory components of a finding: a criterion and an evidence-based condition. The structure widely accepted in the internal audit profession extends this to five elements:
| Element | The question it answers | Mandatory? |
|---|---|---|
| Criterion | What should have been? | Yes |
| Status | What was found? With which evidence? | Yes |
| Why | Why did the gap arise? | Recommended; critical for corrective action |
| Impact | Which risk does this gap create? | Recommended; the basis for rating |
| Recommendation | What could be done? | Optional; watch the consulting boundary |
The gap between criterion and condition is the finding itself. Cause and effect turn the finding from a mere “observation” into information management can decide on. Without a cause, the corrective action goes to the wrong place. Without an effect, the rating is arbitrary.
Criterion: which requirement is it based on?
Every finding must rest on a written requirement the auditee can be shown. Sources of criteria:
- A clause or control of the standard (such as ISO/IEC 27001 clause 9.2, Annex A 5.18)
- The organization’s own policies, procedures and work instructions
- Legal and contractual obligations (KVKK, the security annex of a customer contract)
- Targets or thresholds set by the organization itself
When writing the criterion, cite the most specific source possible. Instead of saying “contrary to ISO 27001 Annex A 5.18”, saying “Access Control Procedure (rev. 3) clause 4.2: ‘Privileged accounts are reviewed by the system owner every three months’” narrows the auditee’s room to dispute and makes the corrective action clear. If the organization’s own procedure sets a stricter requirement than the standard, the finding rests on that procedure. A situation you may come across: the procedure was copied from a consultant’s template and contains frequencies nobody can meet. The finding still rests on that procedure; the fix is to make either the practice or the procedure realistic.
Condition and evidence reference
The condition is a factual description of what the auditor saw. It contains no interpretation, adjectives or generalization. A good condition statement includes:
- Scope and population: Which system, which period, how many records?
- Test and result: How many samples were reviewed, and in how many was a deviation found?
- Evidence reference: Working paper number, report name, screenshot code.
Example: “Review records for the 14 privileged accounts in the ERP system were requested for the first two quarters of 2026. A review record exists for the first quarter (WP-07/1); no record was presented showing that the review for the second quarter was carried out.” The number, period, evidence reference and what’s missing are clear. There’s nothing left to dispute. How the sample was built directly affects the finding’s credibility; for details, see the sampling method in internal audit guide.
Cause and effect analysis
Why
The cause is the condition that led to the gap and usually comes out of the interview with the auditee. The auditor doesn’t have to do a definitive root cause analysis; that’s the auditee’s job in the corrective action process. But writing the observed cause makes the corrective action focus on the right place. In the example above, the cause might be: “After the system owner carrying out the review changed roles in March, the task wasn’t handed over to the new owner; the review schedule is kept in a personal calendar.” This sentence shows the problem stems from the process, not the person. Most recurring findings in the field come out this way: the control is tied to one person’s calendar, and when the person goes, so does the control. Carrying out root cause analysis is covered in the root cause analysis in corrective action guide.
Impact
The effect describes in business terms which risk the gap creates: “Privileged accounts that aren’t reviewed create the risk that people who change roles or leave keep the ability to change financial records in the ERP.” Don’t exaggerate the effect or write it as harm that has already happened. “Has caused a data leak” can’t be written without evidence; “could cause” must be supported by a concrete route. Speculation weakens the finding.
Weak and corrected finding examples
| Problem | Weak wording | Corrected wording |
|---|---|---|
| Vague | “The backup process is inadequate.” | “Backup Procedure clause 5 requires quarterly restore testing; no restore test record for 2026 was presented.” |
| Targets a person | “The system administrator makes changes without approval.” | “In 3 of the 25 changes reviewed (WP-12) there is no CAB approval record; all three were opened as emergency changes.” |
| No evidence | “Users appear to be sharing their passwords.” | Without evidence, no finding is written; if needed, it’s noted as a topic to test in the next audit. |
| Generalizing | “No server keeps logs.” | “On 2 of the 6 servers sampled, the security event log is not forwarded to the central log system (WP-15).” |
| Imposes a solution | “Brand X PAM product should be purchased.” | “It is recommended that a method be set for recording privileged account use.” |
The common principle: a finding must be concrete enough that anyone reading it can look at the same evidence as the auditor and reach the same conclusion. Adjectives (“inadequate”, “weak”, “serious”) belong in the rating, not in the finding text.
An example using all five elements together
Criterion: Access Control Procedure (rev. 3) clause 4.2 requires privileged accounts to be reviewed by the system owner every three months.
Condition: Review records for the 14 privileged accounts in the ERP were requested for the first two quarters of 2026; the first-quarter record exists (WP-07/1), the second-quarter record could not be presented.
Cause: After the owner carrying out the task changed roles in March, the work was not handed over; the schedule is kept in a personal calendar.
Effect: The risk of people who changed roles or left keeping the ability to change financial records went unmonitored for one quarter.
Recommendation: Assigning the review task to a role rather than a person and scheduling it in a corporate tracking tool could be considered.
In this form, the manager reading the report understands the finding in one paragraph, and the corrective action owner sees what to fix. No long explanation is needed.
Rating and agreement
Rating scale
Internal audit findings are rated so that management can prioritize. The major/minor distinction in certification audits serves a different purpose; for a comparison of the two classifications, see the major vs minor nonconformity guide. An example internal audit scale:
| Rating | Definition | Expected response |
|---|---|---|
| High | No control, or it systematically doesn’t work; a significant risk is exposed | Quick action, reporting to top management |
| Intermediate | The control exists but isn’t operated consistently | Planned corrective action, specific target date |
| Low | Isolated deviation, limited effect | Correction at process owner level |
| Observation / opportunity for improvement | Not a nonconformity, an improvement suggestion | Optional consideration |
The rating must be based on the finding’s effect section and the organization’s risk acceptance criteria, not on the auditor’s general impression. If one auditor calls the same deviation “high” and another “low”, the scale definition is incomplete.
Agreeing with the auditee
Agreement doesn’t mean the auditee accepts the finding, but that there’s consensus on the accuracy of the facts. Recommended steps:
- Confirmation on site: Show the deviation to the auditee as soon as you find it; if evidence is missing, give them time to present it. A finding heard for the first time at the closing meeting is met with objections.
- Sharing drafts: Send draft findings to the process owner before the closing meeting.
- Closing meeting: Each finding is read together with its criterion and evidence. If there’s a factual error, it’s corrected; if there’s a difference of interpretation, it’s recorded.
- Management response: The auditee states an action, owner and target date for each finding. The corrective action process under ISO/IEC 27001 clause 10.2 starts here.
- Recording disagreement: If the auditee disagrees with a finding, their view is added to the report. The finding isn’t removed; the final decision belongs to the person responsible for the audit program or to top management.
Pre-publication checklist
- Every finding has a specific criterion reference
- The condition statement includes the period, population, sample and number of deviations
- Every finding is linked to at least one working paper or evidence code
- No personal names or accusatory wording in the text
- The effect isn’t written as harm that has already happened
- The rating is given according to the defined scale
- The auditee’s response or dissenting view has been added
Carrying out internal audits, including writing findings, and following up corrective actions is the subject of our ISO 27001 internal audit and corrective action management service; assessing control effectiveness is the subject of our IT security audit service. For the definition of an audit finding and audit principles, ISO 19011:2018 is the primary source.
Frequently asked questions
Is it mandatory to write a recommendation for every finding?
No. Criterion and condition are the mandatory elements of a finding; a recommendation is optional. If a recommendation is written, it shouldn’t impose a specific product or solution but show in which direction the problem should be solved. Choosing and implementing the solution is the auditee’s responsibility.
If the auditee disputes a finding, is the finding removed?
If there’s a factual error, that is, if new evidence presented changes the condition, the finding is corrected or removed. If there’s only a difference of interpretation, the finding stays and the auditee’s view is added to the report; the final decision is made by the person responsible for the audit program or by top management.
Should the auditor or the auditee write the cause section?
The auditor writes the cause they observed from interviews and evidence. Detailed root cause analysis is the auditee’s responsibility in the corrective action process. The cause the auditor writes gives this analysis a starting point.
Can a major and minor classification be used for internal audit findings?
It can, but these classes are defined in the context of certification audits. In internal audit, a risk-based scale (high, medium, low, observation) is usually more useful so management can prioritize. Whichever scale is chosen must be defined in the audit procedure.