An internal product safety review routine: the calendar that keeps safety work current

Good intentions fade under sales pressure, staff absence, and busy seasons, while a routine survives them because it has a date, an owner, and an agenda. A business that reviews complaints only when someone remembers will miss patterns that a monthly review would catch. A business that reviews files only before an audit will find gaps an authority finds first. Routine is what turns safety knowledge into safety practice, week after week.

A routine also creates a record of diligence. A calendar of reviews, with notes and actions, shows that the business monitored its products continuously, not merely at launch. That record matters to authorities, to partners, and to the business itself, because it proves the safety system was alive between launches, not dormant.

A complaint review should look at all complaints since the last review, grouped by product and by symptom, asking whether any pattern suggests a safety fault, whether any single complaint is serious enough to act on alone, and whether any complaint reveals a labelling or instruction gap. The review should record its conclusions, even where the conclusion is no safety action, because a recorded no action, with reasons, is a decision, while an unrecorded complaint is merely a message that was read and forgotten.

Complaint review needs a threshold for escalation agreed in advance, so a pattern does not wait for a meeting to be recognised. Illustration: a seller of kettles escalates any second complaint about the same fault in a single model to a same week assessment, rather than waiting for the monthly review, and catches a handle fault early.

A file review should sample product files and check them against the documentation mistakes that recur: scope matching, link completeness, version currency, evidence precision, and retrievability. It should focus on files that changed since the last review, because changed files are where stale versions and broken links appear, while unchanged files can be sampled less often. The review should produce fixes with owners and dates, not merely observations.

File review should also check that complaint review actions were filed: where a complaint led to an assessment update or a test, the update should be in the file, not only in a meeting note. That linkage, between review and file, is what makes reviews more than conversations; it makes them changes to the product's safety record.

A supplier review should ask whether suppliers have notified changes, whether incoming inspection results remain steady, whether supplier documents are current, and whether any supplier has generated complaints or test failures. It should also ask whether any supplier relationship has drifted, such as a supplier quietly changing a sub supplier or a material, without notification. Drift discovered in review is correctable; drift discovered through a fault is a crisis.

Supplier review should be proportionate to risk. A supplier of a safety critical component deserves a deeper review than a supplier of packaging, and a supplier with a history of lapses deserves more attention than a new supplier with a clean record so far. Proportionate review focuses effort where failure would hurt most, which is the only way a small team can review a whole supply base.

An alert review should cover Safety Gate alerts for the business's product types since the last review, plus any authority priorities or market changes that affect its range. Each relevant alert should generate a decision: no action, check, or act, with the check or action logged. An alert review that merely lists alerts without decisions is reading, not reviewing, and it will not prevent the next fault.

Market review should also cover the business's own range: products that have been on sale a long time without review, products whose sales have grown sharply, and products sold into new markets or channels. Growth and age both change risk, because a product selling ten times more units creates ten times more exposure, and a product unchanged for years may no longer match current expectations.

A routine sticks when it is calendared, owned, briefed, and followed up. Calendared means dates in a shared calendar, not intentions in a document. Owned means a named person, with a deputy, for each review type. Briefed means the team knows what each review covers and what it produces. Followed up means actions from one review are checked at the next, so unfinished actions do not vanish between meetings.

The routine should also be reviewed itself, periodically, asking whether its rhythms still fit the business, whether any review type is consistently skipped, and whether any new risk needs a new review. A routine that adapts to the business will survive; a routine that fossilises will be abandoned, and abandonment will be discovered only when a pattern goes unnoticed for months.