1. Visible growth or a sudden sour change is a stop
If you can see colonies, coloured spots, fuzz, or a slime film that was not in the retain, the sample is no longer a texture problem. Cap it. Isolate it. Do not stir ‘to check.’ Do not taste. Do not put it back on skin to ‘see if it stings.’ A sharp sour or cheesy odour that the retain does not have is the same class of stop even before you see growth: you are no longer comparing viscosity. Cosmetic QC can say: this unit is out of evaluation. It cannot say what organism it is, and it must not offer medical advice.
2. Odour is a flag, not a species
Sour, acidic, yeasty, putrid, and ‘like old gym bag’ get used as if they were identifications. They are not. They are reasons to stop casual handling. A rancid fatty note is a different flag — paint, cardboard, old nuts — and belongs to the oil-oxidation diagnostic, not to this triage. If both a sour note and a paint note are present, you still stop the unit; you do not average the poems. Write what you smelled in plain language next to the retain, then move the suspect jar off the tasting bench.
3. Water on top is not automatically a culture
Syneresis and a physical split can throw a clear watery layer without growth. That is still a failed emulsion, and it has its own page: timestamp the split, check the emulsifier pair, heat history, and oil load. The difference here is the rest of the picture. A clean watery ring on an otherwise odour-stable retain is a physical job. A watery ring plus sour odour plus spots is a spoilage stop that you do not ‘fix’ by re-homogenising. Browse the emulsifier category when the cream only wept and the retain still smells like day one. A documented lot such as Olivem 1000 does not rescue a jar that already shows growth.
4. The preservative file is not a sniff-test alibi
A cream can list a famous preservative and still fail because the pH walked, the dose was theatre, water activity was high, or fingers went in the jar. Walk the preservatives and antioxidants aisle to rebuild the next batch — not to excuse this one. They do not let you declare a sour prototype ‘probably fine because Phenoxyethanol is on the INCI.’ If the brief used Cosgard preservative in a water mock-up and the cream later sat at another pH, that is a window problem — already published — not a reason to keep evaluating a jar with growth.
5. Rancid oil is a third fork, not a poetic synonym
Carrier-oil rancidity is peroxide and odour of the fat, often without a watery serum ring and without colonies. If the cream smells like old oil and the water phase looks intact, step sideways to that diagnostic and to the oil you actually weighed. If the cream weeps and smells sour, do not call it rancid to avoid a microbiological conversation. Wrong fork, wrong next test.
6. What the lab book needs before anyone argues
Date, batch, pack, retain comparison, pH if you can measure without contaminating more of the jar, visual description, odour versus retain, and whether evaluation is stopped. Photograph in ordinary light. Do not open a obviously grown sample to ‘get a better pH.’ A challenge test is a planned method on a representative mass, not a rescue of a spoiled tester. If the line must continue, that is a new batch with a reviewed preservative system — not a stirred-out colony.
7. What you may tell the workshop — and what you may not
You may say: this prototype is out of sensory evaluation; do not apply it; keep the retain sealed. You may say the failure looks like spoilage red flags rather than a clean split or a rancid oil. You may not name a pathogen. You may not say ‘a bit of essential oil will save it.’ You may not publish a home challenge-test kit. Personal symptoms after exposure are a clinician’s job, not a formula note.