ISO 22000 clause 9.1.1: define measures people can interpret and use

What clause 9.1.1 requires in practice

Clause 9.1.1 turns measurement into a set of explicit choices. The organisation determines what needs to be monitored and measured, and which methods of monitoring, measurement, analysis and evaluation will produce valid results. It sets when the monitoring and measuring is carried out, when the results are analysed and evaluated, and who does that analysing and evaluating.

Two obligations sit alongside these choices. The organisation keeps suitable documented information as evidence of the results, and it assesses how well the FSMS performs as well as how effective it is. Notice the wording on validity: a method that gives unreliable numbers does not meet the intent, however regular the reporting. The standard does not prescribe indicators, targets or a dashboard. A site might choose repeat-failure rates, response times to deviations or coverage of verification checks. Each must be tied to something the system is trying to achieve. The essential output is a clear plan for measurement together with the results that follow from it, available for later analysis under 9.1.2.

Start with the question the measure should answer

Identify the performance or effectiveness question. For example, the organisation may need to understand whether a recurring control weakness has reduced or whether an objective is being achieved. Select information that can answer that question.

Define the measure sufficiently for consistent use. Specify the relevant population, period, source and meaning where needed. Two sites cannot be compared fairly if one counts events and another counts affected lots under the same heading.

Avoid choosing a measure solely because the data is easy to obtain. Convenient activity counts can support evaluation, but they may not establish the outcome management needs to understand.

Choose valid methods and useful timing

Determine appropriate methods for monitoring, measurement, analysis and evaluation. Consider the reliability of the source, relevant sampling or coverage and any limitations. A calculation cannot repair data that does not represent the intended question.

Decide when information is collected and when it is analysed. These timings may differ. Operational product decisions may require prompt action, while a broader trend can be evaluated periodically.

Do not let periodic reporting delay immediate response under product and process nonconformity controls. Performance evaluation complements those decisions.

Assign a capable evaluator

Identify who will analyse and evaluate results. That person needs access to relevant context and suitable competence. A report owner who can calculate a percentage may still need technical input to interpret its food safety significance.

Define how conclusions reach people who can act. A measure with no decision owner can persist for years without affecting the FSMS. Retain appropriate documented information as evidence of results.

Review the measurement arrangement when the operation changes. A new product mix, customer route or recording system can alter comparability with previous periods.

Why clear measure definitions protect decisions

Numbers often cause arguments when nobody agrees what they mean. One depot counts every stock error and another only those found after dispatch; one plant counts a deviation per shift and another per lot. Management then compares sites unfairly, rewards the wrong behaviour or misses a real problem. Defining what is measured, how and by whom prevents that confusion and makes the later analysis worth reading. It also protects staff, since targets set on poorly defined measures tend to push people towards under-reporting. For customers and assessors, a measurement plan that makes sense, with owners who can explain the results, shows a system that is actually managed rather than merely documented.

How to implement clause 9.1.1

Create a short measurement plan that anyone in the food safety team could explain to an auditor. Keep it small enough to maintain, and let every measure earn its place by supporting a decision.

1. State the performance question behind each measure, and drop any measure that nobody uses to decide anything.
2. Write each definition precisely, including the denominator, the data source and the period, and test it with two different sites.
3. Choose methods that suit the data: sampling, counts, observations or trend charts, noting the limits of each.
4. Set separate timings for collection and for analysis, and say which results trigger immediate action instead of waiting.
5. Name the person who evaluates each result and the route by which conclusions reach the people able to act.
6. Store results so that period and scope remain visible, and record any change to a definition with the reason.
7. Revisit the plan when products, customers or recording systems change, to keep comparisons accurate.

Illustrative food-production example

Illustrative food-logistics example

A distributor compares stock-status errors across depots. It discovers that one depot records every correction while another records only errors discovered after dispatch. The team aligns definitions before using the figures to compare performance.

The review also identifies when results should be evaluated and who can address the underlying process. A common chart format alone would not make the data comparable.

Evidence and common mistakes