There’s a wall in almost every certified facility with the quality policy framed on it, and near the policy, the year’s quality objectives. “Improve customer satisfaction. Reduce defects. Enhance on-time delivery.” By March, nobody has looked at them. By the September management review, someone pulls together whatever numbers exist and declares the objectives “on track.” By December, they’re renewed verbatim for another year of being decorative.
Clause 6.2 was not written to produce wall art. Read closely, it’s one of the more demanding clauses in the standard, and one of the most useful when it’s actually run.
What the clause requires, and the word everyone skips
Clause 6.2.1 says quality objectives must be consistent with the quality policy, measurable, monitored, communicated, updated as appropriate, and relevant to product conformity and customer satisfaction. Then 6.2.2 adds the part that separates real objectives from posters: when planning how to achieve them, you determine what will be done, what resources are required, who is responsible, when it will be completed, and how results will be evaluated.
That second list is a project plan. The standard is saying an objective isn’t a wish plus a metric; it’s a wish, a metric, a plan, an owner, a deadline, and an evaluation method. “Reduce defects” fails five of those six. “Reduce final-inspection reject rate on the coating line from 4.1% to under 2.5% by October 31, via the fixture redesign in CAPA-2026-014, owner J. Marsh, reviewed monthly at production meeting” passes all of them.
The three ways objectives go ornamental
Objectives without denominators. “Improve on-time delivery” can’t fail, which means it can’t succeed either. If the objective doesn’t contain a number, a baseline, and a date, it isn’t measurable in the sense 6.2.1(b) means; it’s aspirational in the sense that reads well at kickoff meetings.
Objectives without mechanisms. A target with no 6.2.2 plan behind it is a forecast, not an objective. If the reject rate is supposed to fall from 4.1% to 2.5%, something has to cause that. Which process changes? Funded by what? If nobody can name the mechanism, the honest objective is “hope defects decline,” and hope is not a clause 6.2.2 resource.
Objectives nobody meets about. Clause 6.2.1 requires monitoring, and clause 9.3.2 puts “the extent to which quality objectives have been met” on the management review agenda by name. An objective reviewed once a year isn’t monitored; it’s exhumed. Monitoring means the metric has a home in a recurring meeting where a human is expected to explain the trend, and where a bad trend triggers a decision rather than a font change.
Cascading without the theater
The standard requires objectives “at relevant functions, levels and processes.” Some companies read this as a mandate to give every department its own objectives page, which produces the corporate version of the ornamental register: eleven departments, forty-four objectives, zero connections.
The better reading is a chain. The organization-level objective (cut cost of poor quality 20%) decomposes into the two or three process-level moves that could actually deliver it (halve coating-line rejects; cut supplier-caused receiving NCRs by a third), and each of those carries its own 6.2.2 plan. A department with no line of sight to a top-level objective doesn’t need a decorative one; it needs to be left alone to run its process. Fewer objectives, genuinely connected, will beat comprehensive coverage in every audit we’ve ever conducted.
The evidence trail auditors actually walk
An experienced auditor rarely starts at the objectives page. They start at management review minutes and walk backward: here’s the objective, show me the monitoring data, show me the plan it came from, show me what happened when the trend went the wrong way in May. The finding isn’t usually “objectives are not measurable.” It’s the gap between the objective and everything around it: the metric that appears in the objectives document but no meeting minutes, the action plan that exists only in the quality manager’s memory, the mid-year update nobody recorded.
This is, once again, a linkage problem. When the objective, its plan, its metric, its owner, and the reviews that discussed it are one connected record, the clause 9.3 evidence writes itself and a stalled objective becomes visible in weeks. When the objective lives on a poster, the plan in a project file, the metric in a dashboard someone built in 2024, and the reviews in slide decks, the same facts exist and prove nothing, because assembling them takes an afternoon nobody has.
A fair test for this year’s list
Take your current objectives and score each against five questions. Does it have a number, a baseline, and a date? Can someone name the mechanism expected to move it? Does one person own it? Has a meeting discussed its trend in the last ninety days? And if it’s off track, did anything change as a result?
Any objective that fails three or more isn’t an objective yet. Either finish it, give it a plan, an owner, and a monitoring home, or delete it and let the wall hold the policy alone. A short list the organization is actually driving satisfies clause 6.2 completely. A long list it’s ignoring satisfies nothing, however nicely it’s framed.