What to check before you fill
Updated
Once the product is in jars it is expensive to reject and impossible to adjust. The checks that matter take five minutes and happen while the batch is still in one vessel.
Five minutes that save a batch
The economics are simple. A batch rejected in the mixing vessel costs the ingredients. The same batch rejected after filling costs the ingredients, the packaging, the labels and the time, and a batch that is not rejected at all costs considerably more than that.
The checks
Appearance. Color, uniformity, any visible separation or graininess.
Odor. Against the retained sample from the last good batch, not against memory.
pH. Measured on the cooled finished product, against the specified range.
Viscosity. At a defined temperature, against the specified range, remembering that structure continues to build for hours after manufacture.
Those four are the attributes a stability study tracks, which is not a coincidence: they are the ones that move first when something is wrong.
Against what
Against written specifications with ranges, and against a retained sample. Both halves matter: the numbers catch drift you cannot see, and the retained sample catches the things no number describes.
Keep retained samples sealed and cool, labeled with the batch code, for at least the product's shelf life.
Passing and failing
Record the result whether it passes or not. A log of every batch's pH and viscosity is how a slow drift becomes visible, and a log with only the good batches in it is not a record.
If something is out of range, investigate rather than adjust and proceed. An out-of-range pH is a symptom, and the cause is usually a raw material lot or a process deviation that will recur next month if nobody looks for it now.