ISO 9001 clause 9.1.3: analyse quality data and evaluate results

What clause 9.1.3 requires in practice

Clause 9.1.3 requires the organisation to analyse and evaluate the relevant data and information that monitoring and measurement have produced. The analysis is then used to judge eight matters: whether products and services conform; how satisfied customers are; how well the quality management system performs and how effective it is; whether planning was carried out effectively; whether the actions taken on risks worked; whether the actions taken on opportunities worked; how external providers perform; and whether the system needs improvement. The standard's note mentions statistical techniques as one possible method, which means they are an option and not an obligation. The key idea is judgement: the clause is satisfied when someone has looked at the information, drawn reasoned conclusions on each of these matters and made the conclusions available to those who decide. Collecting and displaying results alone does not meet the intent.

Evaluate the range of results that matters

The analysis must inform evaluation of product and service conformity, customer satisfaction and QMS performance and effectiveness. It must also address how effectively planning was implemented, the performance of external providers and the need for QMS improvement. These subjects are connected, but one measure rarely answers all of them.

The 2026 clause separately identifies effectiveness of actions taken to address risks and effectiveness of actions taken to address opportunities. Preserve that distinction. An action to reduce delivery failure should be evaluated against its risk-related purpose; an action intended to improve capacity or customer value needs evidence of the intended benefit and relevant effects.

The clause does not require a particular statistical technique. Select a method that suits the data and question. A simple comparison may be sufficient for one decision; another may need segmentation, trend analysis or a more specialised method. The method should help explain the result, not make the report appear more technical.

Why clause 9.1.3 matters to the business

Data does not improve anything until someone interprets it. This clause is where numbers turn into understanding: an apparent gain in one area may be caused by a change in workload, and a flat trend may hide improvement in one customer group and decline in another. Careful evaluation protects the business from false comfort and from overreacting to noise. It also gives leaders a defensible basis for the choices they make about supplier changes, staffing and investment, and it gives auditors confidence that claims about effectiveness rest on analysis, not optimism. Because the same evaluation feeds management review, weak analysis weakens every later decision.

How to implement clause 9.1.3

Set up evaluation as a repeatable habit that answers the clause's eight questions. Keep the discussion accurate by inviting people who dislike the headline conclusion to challenge it, because the useful judgements often emerge from the disagreement and not from the first reading of the chart.

1. Create a simple checklist of the matters the clause names, and note which data source speaks to each.
2. Schedule the evaluation shortly after the data is available, with the people who understand the processes involved in the room.
3. Look at results over time and by segment, and test whether a change might be explained by something other than the intended cause.
4. Judge actions on risks and opportunities by what changed, not by whether they were completed.
5. Choose techniques that fit the data, using simple charts and comparisons unless the situation justifies statistics.
6. Write brief conclusions for each matter, including where the evidence is inconclusive.
7. Send the conclusions to the people responsible for improvement and management review, and note the decisions they prompt.

Illustrative BPO example: apparent improvement after a workload change

Illustrative construction example: an opportunity with mixed results

A contractor trials prefabricated service modules to reduce installation time. Site records show faster installation, but interface rework increases where design coordination was incomplete. The review considers both results when evaluating the opportunity action. Completion of the pilot is treated as an activity, while the value of the changed method remains a separate question.

The contractor identifies the project conditions in which the approach is beneficial and strengthens coordination before broader use. The decision links the time information, rework findings and relevant supplier performance. The example does not prescribe prefabrication as a preferred method; it demonstrates evaluation of a chosen opportunity against its intended outcome and effects.

Move from data to a supported judgement

Clarify the period, population and basis of comparison. Check whether definitions or processes changed during that period. A rate can improve because the numerator fell, the denominator grew or the work mix changed. Understanding these differences prevents the organisation from rewarding a reporting effect instead of real improvement.

Use relevant qualitative information alongside numbers. Audit findings, customer comments and process observations can explain why a trend occurred. They should be assessed for relevance and reliability, just as numerical data should. Avoid selecting only information that supports an existing preferred conclusion.

Record the conclusion with enough reasoning to make it useful later. State what the analysis suggests, what remains uncertain and what action or further review is appropriate. The conclusion may be that an action is effective, partly effective or not yet assessable. Uncertainty is more useful than a premature success label.

Strengthen the link between evidence and action

Common nonconformities with clause 9.1.3

Typical findings describe a gap between data held and conclusions drawn.

• Reports present results but there is no record of what was concluded.
• One or more of the matters the clause names, such as provider performance or the effect of risk actions, is never evaluated.
• Action completion is treated as proof of effectiveness without checking the outcome.
• Averages are used where they hide serious problems in a segment or period.
• Evaluation findings are not passed to management review, so decisions are made without them.