Water activity: the one number that runs your shelf life
Moisture content tells you how much water is present. Water activity tells you how much of it matters. Most shelf-life failures trace back to the second number.
Read the article →Every founder wants shelf life proven in six weeks. Every quality manager knows some products only tell the truth after six months. Both positions are defensible, and the choice depends on how your product fails.
What accelerated testing does well
Storing product at elevated temperature speeds up chemical reactions in a broadly predictable way. For oxidation, non-enzymatic browning, vitamin degradation and some texture changes, accelerated data gives a useful early signal. It is the right tool when you need a go/no-go decision on a formulation, or a comparison between two packaging structures.
Where accelerated testing misleads
- Microbial failure. Spoilage organisms do not scale with temperature the way chemical reactions do. A product can look clean at 40 degrees and spoil at ambient because a different organism dominates.
- Phase separation and staling. Physical changes such as fat migration, starch retrogradation and sedimentation follow their own kinetics.
- Packaging interactions. Migration and flavour scalping often need real time at real conditions.
A practical protocol
For most food and beverage products we run both, in parallel, from day one: an accelerated arm at elevated temperature and humidity for early decisions, and a real-time arm at the actual distribution condition for the label claim. The accelerated arm informs development. The real-time arm justifies the date on the pack.
Choose the right conditions for India
Ambient storage in most of India is not 25 degrees. Warehouses in peak summer routinely exceed 40 degrees, and a truck parked at midday is worse. Studies designed around European ambient conditions produce dates that fail in Indian distribution. We normally test at both a controlled ambient and an elevated ambient that reflects the real supply chain.
Decide the endpoint before you start
A shelf-life study is only meaningful if you have defined what failure looks like. That means specifying, in advance, the microbial limits, the peroxide or acid value threshold, the moisture or water activity limit, and the sensory attributes that constitute rejection. Without those, the study becomes a collection of numbers with no decision attached.
How long is enough
Whatever date you intend to print, the real-time study should run at least to that date plus a margin. Launching on partial data is a commercial decision, not a technical one, and it should be made deliberately with a plan to confirm the date before the second production run reaches the market.