Guide · Food safety · UK
A report might be nothing, or the first sign of a failed control. The investigation is how you tell.
When a customer says your food made them ill, the instinct is either to apologise your way out of it or to dismiss it. Both are mistakes. A single report might be nothing, or it might be the first visible sign of a control that has quietly failed. The purpose of a proper investigation is not to win the argument with the customer - it is to find out, honestly and quickly, whether your kitchen has a problem, and to fix it before anyone else is affected.
The first thing an investigation has to overcome is the customer's natural assumption that the meal they ate last is the meal that made them ill. Incubation periods make this unreliable: many foodborne bugs take a day or more to produce symptoms, so the culprit could easily be a meal from before yours. That is not a reason to dismiss the report - it is the reason you investigate with records rather than with your gut. The symptoms and their timing are a clue to what you are looking for, not a verdict on your kitchen.
Gather the facts
An investigation runs on two tracks: what the customer can tell you, and what your own records reveal. Capture both properly and early, while the day in question is still reconstructable.
Record the report factually and without admitting liability or blaming a dish before anything is established - an early admission can be wrong and unhelpful in equal measure. Ask the customer, tactfully, to keep any remaining suspect food chilled rather than binning it. Then turn inward and pull every record that touches the meal they describe. This is where a kitchen that keeps honest, contemporaneous logs has an enormous advantage: it can actually reconstruct the day.
Find the cause
The heart of the investigation is root-cause analysis: take the implicated dish and walk it back through every step of its journey, asking at each one whether a control could have failed. Was the core cooking temperature reached and logged? Was cooling done fast enough, or did something sit in the danger zone? Was hot-holding above the safe threshold? Could raw and ready-to-eat have crossed? Was a supplier ingredient itself suspect? Traceability matters here - being able to identify the batch and its supplier lets you tell the difference between a one-off handling slip and an ingredient problem that affects far more than one plate.
You are looking for a mechanism, not a scapegoat. If the cooling log shows a large batch that missed its target, you have a plausible cause and a clear fix; if every control held and is evidenced, that genuinely points away from your kitchen. Being able to carry out supplier due diligence and produce a full traceability trail in minutes is often what separates a contained answer from a blanket recall.
Act and escalate
Whatever the investigation finds, document it and act on it. If you identify a failed control, correct it, retrain if needed, and record the change - that record is your evidence of a system that learns. Know the escalation triggers too: you have a legal duty to inform the local authority if you have reason to believe food you placed on the market is unsafe, and two or more illness reports from separate bookings should be treated as a possible outbreak and taken to your environmental health officer promptly. Cooperate fully; an operator who investigates properly and shares findings is in a far stronger position than one who went quiet.
The investigation and the customer conversation are two different jobs, and both matter. The internal root-cause work is what protects the next customer; handling the person in front of you with care is what protects the business. Doing the second well is the art of managing a food safety incident without losing customers, while the first is closely related to how you handle a suspected food poisoning complaint at the point it lands.
Questions
Record the report factually - symptoms, when they started, how long after eating, who else was affected, whether a doctor was seen, and the exact meal, date and time - without admitting liability or blaming a dish. Ask the customer to keep any suspect food chilled. Then pull your own records for that day and begin an internal investigation.
Not necessarily. Many foodborne illnesses have an incubation period of a day or more, so the meal eaten just before symptoms began is often not the cause. Use the symptoms and their timing as a clue to guide your investigation of the records, not as a verdict.
Taking the implicated dish and tracing it back through every step - cooking temperature, cooling, hot-holding, cross-contamination, and the supplier ingredient - to find where a control could have failed. You are looking for a mechanism to fix, not a person to blame. Traceability lets you distinguish a one-off slip from an ingredient problem.
You have a legal duty to inform the local authority if you have reason to believe food you placed on the market is unsafe. Two or more reports from separate bookings should be treated as a possible outbreak and taken to your EHO promptly. Cooperate fully and share your findings.
Document what you found and act on it. If a control failed, correct it, retrain staff where needed, and record the change - that record demonstrates a system that learns and supports your confidence-in-management score at the next inspection.
Phoenix Duct Clean · by the numbers
Cross-contamination and hygiene sit behind many illness investigations. A certificated deep clean cuts those root causes and evidences your control.