Post-market monitoring and complaints: watching products after sale so a scattered worry becomes a decision in time

A product that passed its launch tests has not finished its safety story, because production drifts, suppliers change, users find new uses, and materials age. A heater that was safe at launch may develop a fault when a new switch supplier is used. A toy that was safe for its first batch may shed parts when a new adhesive is introduced. A kettle that was safe in testing may be used in ways testing never pictured, such as in a damp garage or by a child. Monitoring after sale is how a seller learns those later chapters, instead of learning them from a serious incident or an authority notice.

Monitoring also protects the seller commercially: a pattern caught early can be fixed with a supplier change or an instruction update, while a pattern caught late may require a recall, a retailer dispute and lasting reputational harm. The cost difference between an early fix and a late recall is usually large, and the early fix is usually simpler, since it addresses a cause before it spreads across batches. A monitoring routine is therefore both a safety duty and a business habit, and it should be presented that way to teams who might otherwise see it as paperwork.

Monitoring signals come from many places, and a good routine gathers them all: customer complaints and enquiries, returns and refunds, marketplace reviews and questions, social media mentions, retailer feedback, test laboratory results, supplier notices, and regulatory alerts including Safety Gate. Each source has its own bias: complaints overrepresent angry customers, reviews overrepresent extremes, and returns overrepresent visible faults, while silent failures, such as a safety feature that quietly stops working, may appear in no source at all. Gathering multiple sources, and comparing them, corrects for those biases better than any single source can.

A complaint is a single report from one customer about one product experience. A signal is a pattern across reports that suggests a common cause, such as several reports sharing a batch, a part, a use or a failure description. Noise is the background of unrelated reports, vague descriptions and one off faults that do not form a pattern. Distinguishing signal from noise requires grouping reports by product, batch, part and failure words, and looking for repetition that is unlikely to be coincidence. A single report of a serious harm, such as a burn or a swallowed part, deserves individual attention even without a pattern, since severity can outweigh repetition.

Illustration: a seller of electric blankets received 40 complaints in a season, most about warmth or size, but six mentioned a controller that grew hot and smelled, all from one batch. The six, not the forty, formed the signal, and the seller investigated that batch and its controller supplier. The complaint count is illustrative, and the lesson is to group by failure words and batch, rather than counting complaints in total, since a total hides the small, serious pattern inside a large, harmless one.

A proportionate routine for a small seller can be simple: one log for complaints and signals, one weekly review of new entries, one monthly pattern check across products, and one named person responsible for each step. A larger seller may need a database, a triage team and formal escalation thresholds, but the logic is the same: collect consistently, group by product and batch, review regularly, escalate promptly when a pattern or a serious report appears, and record decisions. The routine should be written down in a page or two, so a new team member can follow it and an authority can see it was followed.

Every monitoring decision, whether to investigate, to watch, to contact a supplier, or to take corrective action, should be recorded with the signal, the evidence, the reasoning, the decision, the person who made it, and the date. That record lets a later reviewer see why a signal was watched rather than acted on, or why an action was taken, and it prevents a signal from being forgotten when staff change. Follow up matters equally: a decision to watch a signal should carry a review date, a decision to contact a supplier should carry a reply deadline, and a decision to act should carry a completion check, so decisions do not drift into inaction.

This month, start with one product and one log: gather its complaints, returns and reviews into a single sheet, group them by batch and failure words, and review the sheet with a colleague. Note any pattern, any serious report, and any supplier question, and record the decisions made. Repeat for the next product, and then make the log and the review a weekly habit. A monitoring routine started small, with one product done well, grows more reliably than a grand system announced for all products at once and never used.