Stability testing protocols are the difference between a shelf life that is declared and a shelf life that is known. Every cosmetic carries a date; the date should come from data, and the data comes from a matrix of temperatures, times and tests that has to be designed before the first batch rather than after the launch.
1. The temperature matrix
Three conditions cover most products: 25 degrees Celsius at 60 percent relative humidity for real-time, 30 degrees Celsius at 65 percent for a warm-climate market, and 40 degrees Celsius at 75 percent for accelerated. A freeze-thaw cycle set – three cycles between minus 10 and 25 degrees – catches emulsion and pack failures that a warm hold misses. A product sold in a tropical market needs the 30-degree condition as its primary real-time, not the 25-degree one.
2. The pull schedule
Accelerated: 0, 1, 2 and 3 months. Real-time: 0, 3, 6, 12, 18 and 24 months. A 12-month claim with only three months of accelerated data is a prediction, and a buyer will ask for the real-time confirmation. The practical approach is to launch on accelerated plus early real-time data, and to keep the real-time programme running so that the claim can be confirmed before the first anniversary.
3. What to test at each pull
Appearance, odour, pH, viscosity, colour, emulsion integrity, weight loss, microbial count, and the active assay where the product carries one. For a pack, add leakage, pump performance and label adhesion. Testing only appearance and pH at the pulls is the most common shortcut and the one that produces a launch followed by a complaint at month four.
4. The specification
A stability programme is only useful if each parameter has a defined acceptance range before the testing starts. Viscosity plus or minus 15 percent, pH plus or minus 0.3, no visible separation, microbial count within limits, active content above 90 percent of label. A failure against a written specification is a decision; a failure against an unwritten expectation is an argument.
5. Weight loss and the pack
Weight loss is a pack property as much as a formula property. Measure the filled unit at every pull, and set a limit – typically 2 to 3 percent over the shelf life. A product that loses 6 percent loses texture at the surface and concentrates its actives, and the customer experiences it as a product that changed.
6. Compatibility and the interaction
A product in contact with its pack is a two-material system. Test the filled unit inverted for 30 days, then check for stress cracking, discolouration, swelling and label lifting. A fragrance-heavy product in a low-density polyethylene bottle can swell the bottle and change the dose; a product with a high alcohol content can stress-crack a polystyrene cap.
7. The question a buyer actually asks
A distributor will ask two things: what is the shelf life, and what is the evidence. The answer has to be a one-page summary – conditions tested, pulls completed, parameters in specification, and the claim supported. A brand that can produce that page within an hour sells into more channels than one that has to email a laboratory in another time zone.
8. Commercial profile
Accelerated programme, three months; real-time programme, twelve to twenty-four months, run in parallel with production. See the cream range and the microbial guide for the tests that run at each pull.
