Stability Testing for Cosmetics: What Brands Must Know
Stability testing is the point in a project where a promising sample becomes a product that can be sold. It answers three questions at once: does the formula hold together physically, does it stay chemically potent, and does it remain microbiologically safe over the period the customer will own it. Most first-time brands treat it as a formality and most first-time brands are surprised by the result.
What Is Actually Measured
A complete stability programme checks more than whether the emulsion has split. The usual panel of measurements includes:
- Appearance and homogeneity. Colour, gloss, phase separation, grittiness, and for emulsions any sign of creaming at the surface.
- Odour. Assessed by a trained panel, because rancidity and ingredient breakdown often appear as smell before anything visible changes.
- Viscosity and pH. Both can drift without any visible change, and either can move a product outside specification.
- Active content. For products with an active that can degrade, an assay at each time point. Vitamin C, retinol, probiotics and many botanical extracts all fall into this group.
- Microbiological load. Total aerobic count, yeast and mould, and absence of specified organisms.
- Pack compatibility. Weight loss through the pack, deformation, corrosion, colour transfer, and interaction between the product and the closure liner or pump.
Accelerated Testing and Its Limits
The standard accelerated condition is 40 degrees Celsius with 75 per cent relative humidity, held for three months. The working assumption is that this approximates roughly two years at ambient, and it is used to make an early decision about whether a formula is worth pursuing.
The assumption has limits and it is worth knowing them. Reactions do not all accelerate at the same rate, so a formula can pass accelerated testing and fail at ambient, particularly where the failure mechanism is a slow crystal growth or a gradual oxidation. Freeze and thaw cycling, typically three cycles between about 4 and -10 degrees Celsius, and elevated temperature cycling between 4 and 45 degrees Celsius, add information that a constant 40 degrees Celsius does not give.
The correct sequence is accelerated testing to screen, then real time testing at ambient and at a representative warm market temperature to confirm. Real time data takes as long as the shelf life you are claiming, which is exactly why the decision to keep testing has to be made alongside the launch plan rather than after it.
What Failure Looks Like
Failure is rarely dramatic. The common outcomes we see across projects are:
- Viscosity loss of 20 to 40 per cent over three months. Usually a thickener or a structuring system that cannot hold under thermal stress. Often fixable with a different rheology modifier, sometimes not worth fixing.
- pH drift. Frequently caused by an ingredient hydrolysing in water, by a container interaction, or by an unstable buffer. Urea products and acid products are the usual suspects.
- Colour shift. Common with botanical extracts, vanilla and citrus fragrance notes, and any formula containing certain pigments.
- Active loss. The most expensive failure because it usually means a different delivery format, not a minor adjustment.
- Weight loss through the pack. Water diffusing through the container wall or an imperfect seal. It can take the product below its declared weight and, in a small pack, breach nominal quantity rules.
Planning the Programme With the Project
Stability testing has a lead time that has to be built into the launch plan. A three month accelerated programme cannot be compressed, and it sits between the sample approval and the production run. If the product needs real time confirmation, the data will arrive after the first production run, which is normal but has to be understood.
The practical scheduling rule is that stability should start as soon as a formula is frozen and the pack is confirmed, running in parallel with the safety assessment and the artwork rather than after them. Projects that run everything sequentially add three months to the timeline for no benefit.
One cost saving worth knowing: once a base formula has been validated, a variant that only changes fragrance or colour often needs a reduced programme rather than a full one. Ask before assuming the full cost applies to every SKU.
Documentation the Assessor and the Buyer Will Ask For
Stability data feeds directly into the safety assessment and, in the European Union, into the product information file. A retail buyer will typically ask for the report summary, the testing conditions, the batch numbers and the conclusion. Keep the raw data as manufactured by batch, since a repeat production run at a later date should be checkable against the validated reference.
Also keep a retained sample of each batch for the duration of the shelf life plus a margin. It is the only way to investigate a complaint about a specific production run eighteen months later.
Stability Programmes Run Alongside Development
OEM COSMETICS ODM runs accelerated, real time and cycling stability programmes in parallel with formula development on ISO 22716 and GMP certified lines, with pack compatibility and microbiological panels included. Send us your target shelf life and destination market and we will map the testing schedule onto your launch date.
Talk to our team: WhatsApp +86 18709713948 · Email adon@oemcosmeticsodm.com · Website www.oemcosmeticsodm.com