Running a Launch Across Two Fillers Without Splitting Quality

The order outgrows one line, or procurement insists on a second source, and the launch becomes a perfume production split. The bulk is the same, the bottle is the same, the specification is the same — and the product is not, because two fillers bring two crimp tools, two pump lots, two sets of operators and two quality regimes that only look identical on paper.

This article sets out the three ways a perfume production split can be organised, where the difference enters the product, what a single quality system across two sites has to contain, and how batch codes and labels keep a recall traceable.

Table of Contents

• Three ways to split a launch
• Where the difference actually enters
• Split models compared
• One quality system across two sites
• Batch codes, labels and traceability
• Practical guide: qualifying a second filler
• Frequently asked questions
• Conclusion

Three ways to split a launch

In a perfume production split, the first decision is what gets duplicated. Duplicating the fill is cheap and controllable; duplicating the bulk is where formula variance begins.

Split fill from one bulk

The concentrate is blended once, in one vessel, and the finished bulk is transferred to two filling sites. Both sites fill the same liquid, so the fragrance is identical by construction. What varies is the fill, the closure and the handling between the two events. This is the default model for a launch that outgrows capacity, and it is the easiest to defend to a retailer because the bulk lot number is common.

Two bulks, two sites

Each site compounds its own bulk from the same formula and specification. This is what happens when the volume is too large to move economically, or when the two sites are in different regions and import duty makes a single bulk impractical. Formula variance now includes raw material lot differences, weighing accuracy, mixing energy and operator technique. Natural materials vary between lots by nature; a synthetic formula varies less but not zero.

Full duplication

Two independent supply chains: separate bulk, separate components, separate filling. It offers the most resilience and the least consistency, and it is only worth it when the risk being insured against is larger than the cost of divergence — a single-site failure, a customs block, a customer that requires dual sourcing.

Where the difference actually enters

Most splits go wrong in ordinary places, none exotic, all invisible in a purchase order.

The bulk transfer

Moving finished bulk from a blending tank into drums or an intermediate container and then into a second site's holding tank introduces three risks: ethanol loss from an open or poorly sealed container, stratification if the bulk sits without agitation, and contamination from a container that was not dedicated to fragrance. Transfer under nitrogen or in a closed lined IBC reduces the first; a defined agitation time before filling addresses the second; a container log addresses the third.

Crimp, pump and glass

Each site crimps with its own die, set by its own mechanic, checked with its own gauge. The nominal dimension can be identical and the capability different: one site might hold ±0.05 mm while the other drifts with die wear between services. Pumps from different production lots can differ in gasket hardness and dip tube length. Glass from different mould cavities differs in weight and, if decorated, in registration. None of these shows up in a Certificate of Analysis for the concentrate.

Why the same specification produces two results

A specification states a target and a tolerance; it does not state the process capability behind it. Two sites can both be "in specification" while one sits at the centre of the tolerance and the other runs against both limits. When a customer compares two bottles, they are comparing distributions, not specifications.

Split models compared

The last row matters: a split answers a constraint, and the cheapest fix is sometimes to remove it at one site. Extending shifts or adding a second line keeps one quality system and one set of tools.

One quality system across two sites

The goal of a perfume production split is not identical sites; it is one specification, one acceptance method and one decision rule, applied at both.

Batch record and retain samples

Use one batch record template with site-specific fields rather than two documents that drift apart. Each site retains samples from the first, middle and last of every run, plus a bulk retain where it compounds. Store them under the same conditions so a later comparison is meaningful. The batch record should name the bulk lot, the component lots and the crimp die used.

The golden sample and the cross-site panel

Keep a sealed golden sample per site, stored dark at controlled temperature, and replace it on a defined cycle. At least once per launch, run a blind panel across both sites' output against the golden sample, scored on the same attributes. Do the same for colour and clarity. The panel is not there to prove the sites are the same; it is there to detect when they have drifted apart.

Batch codes, labels and traceability

A split launch has to be recallable by site. If a defect appears in one market, the batch code should tell you which site filled it and which bulk lot it came from without opening a spreadsheet.

Use a site prefix in the batch code, and record the mapping in one place both sites can reach. The label must carry the manufacturer identification that the destination market requires; where two sites manufacture, both may need to be identifiable, and the responsible-person details must be correct for each. Requirements differ by market and change, so confirm the current official text before the artwork is signed off rather than copying last year's label. Packaging artwork and label space are covered under packaging formats, and the regulatory side in this note on packaging rules.

Then test the trace: pick a batch code at random and ask each site to produce the bulk lot, component lots, fill date, line and QC results within a stated time. A mock recall that takes days is a finding, not a drill.

Practical guide: qualifying a second filler

Step 1 — Decide what is being duplicated

Fill only, or bulk as well. Write the answer down, because it determines which variance sources are in scope and which controls are mandatory.

Step 2 — Compare capability, not certificates

Ask both sites for fill weight and crimp capability studies on a comparable pack. A certificate says the system exists; a capability figure says what the line actually holds.

Step 3 — Send the same bulk and the same components to both

Run a bridging batch: one bulk lot split between sites, one component lot at each, and compare outputs. This isolates the site from the materials.

Step 4 — Harmonise the gauges and the drawing

One crimp drawing, one gauge type, one calibration schedule. Two sites measuring the same crimp with different gauges are not measuring the same thing.

Step 5 — Set a deviation protocol before the first order

Who decides when a site is out of tolerance, who is told, and what happens to the affected units. A protocol written after the first deviation is a negotiation with a customer attached.

Step 6 — Run the mock recall

Both sites, one batch code, a stated time limit. If the records cannot be produced, the split is not ready for a launch.

Frequently asked questions

What is private label perfume?

A fragrance manufactured for a brand rather than for the plant's own label. The buyer sets the formula and pack; the manufacturer holds the process documentation and releases each batch.

What is a minimum order quantity in perfume?

The smallest run a filler will schedule before changeover costs outweigh the order. Two sites mean two minimums, so a split can double the entry quantity.

How long does a split launch take?

Add qualification and bridging to the normal lead time: component duplication, a cross-site panel and a recall test usually add several weeks before the first batch.

Which is safer: one bulk split or two bulks?

One bulk split, because the fragrance is identical by construction. Two bulks are acceptable when lot-to-lot variation is controlled and bracketed by panel testing, but they carry a variance the buyer has to accept.

Do both fillers need the same certification?

They need what the destination market requires for the product and the process, and the buyer should verify both rather than assume equivalence. The scope of a certificate matters as much as its presence.

How do I compare crimp quality across two sites?

Same drawing, same gauge type, same sample size, same time window, then compare means and spreads. A single measurement proves nothing; distributions show whether one line is drifting.

What goes on the label when two sites fill?

At minimum, enough manufacturer and batch information to identify the filling site. The exact requirement depends on the market, so confirm the current official text before artwork approval.

Can a second filler use the buyer's components?

Yes, and it should where consistency matters. Sending one component lot to both sites removes glass and pump variation from the comparison and makes any remaining difference attributable to the line.

Conclusion

A perfume production split is a variance management problem disguised as a capacity problem. The fragrance can be made identical, but the fill, the closure and the record keeping are duplicated, and that is where a customer will notice the difference — usually in a bottle that leaks or sprays weakly, months after the two sites were signed off as equivalent.

Decide what is duplicated, qualify both sites on capability rather than certificates, bridge them with a common bulk and a common component lot, and prove the trace before the first pallet ships. Our customisation process sequences the bridging run and the pack approval so both sites are qualified against the same freeze.

Discuss your fragrance programme with our team:
WhatsApp: +86 17324762943
Email: adon@customperfumeoem.com
Website: customperfumeoem.com