Brand owners tend to review the finished product certificate and stop there. The number that actually decides whether a shipment is compliant, however, is generated on the filling line, one bottle at a time. Net content is a legal statement, and it is proven by in-process data, not by a final certificate. This article explains what a well-run in-process check programme looks like and which records a buyer should ask for.

1. What in-process control actually covers

In-process control is a set of checks performed during production rather than at the end. On a typical skincare line it includes fill weight, cap torque or crimp dimensions, label position, seal integrity, batch coding and appearance. Every check exists because a specific failure mode is cheaper to catch in line than after palletising.

The frequency matters as much as the check itself. A line running several thousand units an hour needs sampling at fixed intervals, with the interval defined by the process capability rather than by convenience. Ten units every hour is common for fill weight on a stable line; a newly commissioned line should start tighter and relax only after the data supports it.

2. Fill weight: the statistics behind the label

Net content rules generally work on two principles: the average of a sample must meet or exceed the declared quantity, and the number of units below a defined tolerance must stay within a permitted limit. The average is not enough; the spread is what regulators and retailers look at.

In practice this means the target fill is set slightly above the declared volume to absorb normal variation. How much above depends on the filling machine, the viscosity of the product and the container tolerance. A low-viscosity toner fills very consistently; a thick cream with entrained air does not. Setting the target without measuring the spread is how shortfalls reach the market.

What to record

Record the individual weights, not just the average, together with the time, the filling head and the operator. Individual data allows a trend to be spotted; a single average hides a drifting head until the batch is half finished. Where a checkweigher is installed, confirm it is challenged with test weights at the start of the run and after any adjustment.

3. Closures, torque and seal integrity

For pump and spray formats, the closure is part of the product performance. A cap that passes a visual check can still leak in transit, and the resulting complaint looks like a formula problem rather than a capping problem. Torque testing with a calibrated gauge, at the same interval as fill weight, gives a number that can be trended.

Induction seals, sachets and tubes each need their own integrity test. Peel strength for sachets and a squeeze or vacuum test for tubes are quick, destructive and worth the cost of a few units per shift. For pressurised aerosol formats, a water bath check at a defined temperature and dwell time is the standard final gate. Seal and closure checks protect the batch from failures that only appear weeks later.

4. Line clearance and first article approval

Line clearance is the step that prevents the previous product from appearing in the next one. It should remove components, bulk, labels and printed matter from the line, and it should be signed off by a second person. Line clearance is a two-person task, not a formality signed at the end of a shift.

First article approval follows: the first filled unit is checked in full against the specification, including label text, batch code, fill level and closure. Nothing runs at full speed until that unit is approved and retained. This single habit catches most label and coding errors before they become a pallet of rework.

5. Records, trending and what buyers should ask

In-process records should be contemporaneous, legible and traceable to the batch record. A buyer reviewing a batch should be able to see the check frequency, the results, the actions taken when a result was out of limit, and the name of the person who approved the release. Contemporaneous records are the difference between a controlled process and a reconstruction.

Ask for a trend of fill weight averages and ranges over the last ten batches of the same product. A stable range suggests a controlled line; a wandering range suggests worn nozzles or inconsistent product viscosity. Trending reveals maintenance needs before they become deviations.

None of this is exotic. It is the routine discipline that separates a factory able to hold a specification over years from one that performs well on the audit day. Production conditions are described on the workshop page, the formats available are listed under products, and questions about specific lines can be raised through inquiry, the essence spray product page or the news section.

Good in-process control is boring by design. That is exactly the point.