ロゴ

How a European Beauty Manufacturer Scaled a Multi-SKU Cosmetic Packaging Program

Alexandre Wong

Alexandre Wong

Lead Technical Engineer

A European beauty and personal-care manufacturer producing for multiple brand programs consolidated five distinct packaging formats with Jarsking over roughly seven months — glass jars in two capacities, a roll-on applicator bottle, an inverted PET bottle, a sunscreen bottle, and a fine-mist spray bottle — growing from one trial order to more than 1.5 million units across seven stock-keeping units. Individual first orders ranged from roughly 12,000 units on the smallest format to over 900,000 on the largest, a spread of nearly 80 to one inside a single account.

One batch was reported from downstream testing as exceeding an applicable packaging heavy-metal limit. We escalated it to management, spoke to the customer directly, re-specified the decorative material, remade the batch and absorbed part of the rework and freight cost. The program continued, and one format has since been reordered. This is what building a multi-format packaging program looks like — including the part most case studies leave out.


Case at a Glance

Scope
7 SKUs
across five packaging formats
Volume
1.5M+
units in first orders
Timeline
6 orders
across roughly seven months
Outcome
Reordered
one format confirmed; two more expected

Figure 1  Program scope at a glance. Volumes are shown as ranges.

ClientA European beauty and personal-care manufacturer producing for multiple brand programs
ScopeSeven SKUs across five formats, in glass and PET, with pumps, sprayers, flip-top closures, a roll-on applicator and an orifice reducer
VolumeOver 1.5 million units in first orders; ~12,000 to 900,000+ units per SKU
TimelineSix purchase events across roughly seven months, from Q3 2025
Core challengeFive unrelated formats and a near-80× volume spread under one quality and program layer
IncidentOne batch reported non-compliant against a packaging heavy-metal limit; re-specified, remade, reshipped
Verified outcomeOne confirmed reorder placed about seven months after that format's first production run; repeat volumes expected on two others

It Started With Two Glass Jars

The program did not open with a portfolio award. It opened with one purchase order covering two glass jar capacities — 15 ml and 50 ml — in the low tens of thousands of units each.

That is, in our experience, the right way for a manufacturing buyer to start. An order of that size exercises a real production run rather than a sample batch: the mold has to run, the decoration has to hold registration across thousands of pieces, the packing has to survive freight, the paperwork has to arrive complete. It is small enough that failure is recoverable.

Both jars came from an existing structure in our glass packaging library rather than new tooling. The customer selected a form, we produced visuals and samples against it, and the approved sample became the production reference. No new mold was cut, which removed tooling lead time and cost from the customer's first decision.

The Challenge: Why Five Formats Become Five Separate Risks

Here is the problem the rest of the program was solving. A glass jar, a roll-on bottle, an inverted PET bottle, a sunscreen bottle and a fine-mist sprayer are not variations of one thing; they are five manufacturing problems that happen to sit in the same product range.

Glass is a hot-forming process with its own tolerances and annealing behaviour. Injection and blow molding are different processes again, with different shrinkage and wall-thickness constraints. A roll-on ball is an assembled mechanism in prolonged skin contact. A fine-mist sprayer is a dosing component. A sunscreen bottle at high volume is a throughput and repeatability problem.

The coordination load compounds rather than scales.

Sourced separately, each format brings its own supplier, sampling cycle, acceptance standard, compliance paperwork and delivery promise. That is tolerable at two formats. At five, the buyer runs five parallel qualification programs with no single party accountable when one slips — because the formats have to arrive in a usable sequence for filling, not merely arrive.

Jarsking's Response: One Program Layer, Three Manufacturing Routes

Our answer was not to claim every format runs on one line. It does not. It was to keep a single program layer above three manufacturing routes.

Three routes, because seven SKUs do not divide evenly. The glass jars needed hot forming; the PET bottles needed injection and blow molding — different processes, different tolerances, in practice different workshops. And four formats turn on a bought mechanism rather than a container: a pump, a fine-mist sprayer, a flip-top with an orifice reducer, and a roll-on ball and housing. Those are specified and qualified, not molded alongside the bottle.

1
Self-owned production
Where the format sits inside our own glass or plastic capability.
2
Self-owned with outsourced finishing
Where a process such as plating or spraying is specialist work.
3
Vetted partner production
Where a component such as a particular sprayer or closure is better made by a specialist.
▼   ▼   ▼
One program layer — one program manager, one quality standard, one document set, whichever route produced the part.

Figure 2  How each format is routed to production and returned through a single accountability layer.

The map earns its keep when something goes wrong, because the first question is always which party owns the step that failed.

Volume tiers follow the same logic. A 12,000-unit order and a 900,000-unit order are not the same job even when the part looks similar: the small run is dominated by changeover and setup, the large run by cycle time and lot consistency. In this program the two sat about a week apart on the order calendar — precisely what a split supply base handles badly, with each vendor quoting a delivery date in ignorance of the other's sequence.

The route changes; the layer above it does not. One program manager owned the account, one quality standard applied to every route, and one set of documents came back regardless of which workshop made the part. That is the actual product in a multi-format program: not floor space, but one point of accountability across routes the buyer never has to see.

How the Program Grew, Order by Order

The program expanded in six purchase events across roughly seven months, and the sequence is the interesting part.

  • Q3 2025The glass jarsTwo capacities on one order, low tens of thousands of units each. The trial.
  • Q4 2025The roll-on bottleA step change in scale: low hundreds of thousands of units on a single 25 ml roll-on format. This is where the relationship stopped being a trial. A buyer does not place an order that size on an unproven supplier; the glass run had answered the questions that mattered.
  • Late Q4 2025The inverted PET bottleA 50 ml upside-down bottle with a pump — a format that has to sit stably on its closure and evacuate cleanly from the top. Mid tens of thousands of units.
  • Late Q4 2025The sunscreen bottleThe largest single SKU, above 900,000 units on a flat-oval bottle with a flip-top closure and orifice reducer. Placed about a week after the inverted PET order, so two very different volume tiers ran concurrently.
  • Q1 2026A second 50 ml glass jarUnder 70,000 units. The customer returned to glass, this time with confidence built.
  • Q1 2026The fine-mist spray bottleThe smallest order in the program, in the low tens of thousands, on a 60 ml PET bottle with a fine-mist sprayer and a dome overcap.

Figure 3  Purchase events across the first seven months. Red markers show the two orders that changed the scale of the program.

Note the shape of that sequence. It is not a portfolio handed over at once: one format proven, then a much larger commitment, then progressively wider scope. The sequencing was the risk control.

What Each Format Demanded

Glass jars, 15 ml and 50 ml

Existing structure, no new tooling. The work was decoration consistency and dimensional control at the closure interface, so one lid specification fits reliably across a lot.

Roll-on bottle, 25 ml

An assembled applicator in prolonged skin contact. The ball and housing have to rotate freely, seal when static and survive freight without leaking — and the ball's material carries a regulatory dimension we return to below.

Inverted PET bottle, 50 ml

An inverted format stands on its closure, so the closure carries structural load as well as sealing duty. Wall thickness, base geometry and dip-tube length have to be resolved together rather than in sequence.

Sunscreen bottle, 50 ml

Above 900,000 units the engineering question is repeatability, not novelty. A flat-oval bottle with a flip-top and orifice reducer is a well-understood format; holding the neck finish, hinge and reducer fit across a run that size is the task.

Fine-mist spray bottle, 60 ml

A dosing component. Spray pattern and shot weight have to stay in tolerance across the lot, which makes the sprayer supplier and lot-level checks more consequential than the bottle.

When a Batch Came Back Non-Compliant

One batch of packaging that had already been filled downstream was reported as exceeding an applicable packaging heavy-metal limit. This is what caused it, what we did about it, and what changed afterward.

Why it happened

The source was the decorative material — the ink and surface finish applied to the container, not the container itself.

That distinction matters to anyone specifying a decorated pack. The substrate is rarely the whole compliance picture: inks, coatings, metallised finishes and closure components all contribute to what a composition test measures. Decoration is also the layer most easily treated as a cosmetic detail rather than a declared material, and the part of a pack most often left out of a brief — a specification that names the bottle, the closure and the resin, then describes the artwork in visual terms alone.

A container can be specified correctly, molded correctly, and still fail on what was printed onto it.

Timing made it consequential. A non-conformance caught at incoming inspection is a supply problem. This one surfaced after filling, which makes it a commercial problem: product has already absorbed cost and a launch window is already committed.

How we resolved it

The case went to management immediately rather than staying inside day-to-day account communication, and we spoke to the customer directly. That is the step that makes everything after it possible; a result of this kind is not resolved by two account contacts trading emails.

We then re-specified the decorative material rather than re-testing the same specification and hoping for a different reading. The affected batch was remade against the revised specification, retested before shipment, and shipped. Jarsking absorbed part of the rework and freight cost.

01Result reported
02Escalated to management
03Decoration re-specified
04Batch remade
05Retested before shipment
06Shipped

Figure 4  How the affected batch was re-specified, remade and retested before shipment.

How it is controlled now

The revised decoration specification is the standard for that pack, and the approved sample together with its decoration specification — not the container geometry alone — is the reference for every repeat run.

More broadly, decoration belongs on the material declaration alongside the container and the closure, and it is subject to change control, because an ink or coating substitution that looks identical can move a compliance result. Jarsking maintains a heavy-metals screening report covering lead, cadmium, mercury and hexavalent chromium, alongside ISO 9001:2015 quality-management certification and ongoing REACH SVHC screening, listed on our certifications page.

The customer's largest orders in this program were placed in the same period as this batch, and the program has continued since.

The Rule: Four Metals, One Combined Limit

Here is the rule, stated precisely.

Four heavy metals are restricted in packaging and packaging components: lead, cadmium, mercury and hexavalent chromium. The restriction applies to the sum of their concentrations, which must not exceed 100 mg/kg — not to each metal individually. It is set out in Article 5 of Regulation (EU) 2025/40, the Packaging and Packaging Waste Regulation, which applies across the EU from 12 August 2026.

RegimeApplies toSubstance scopeLimitBasis
PPWR, Reg. (EU) 2025/40, Art. 5Packaging and packaging componentsLead, cadmium, mercury, hexavalent chromiumSum not exceeding 100 mg/kgComposition of the material
REACH, Annex XVII, entry 27Articles in direct and prolonged skin contactNickel and nickel compounds0.5 µg/cm² per week releaseEN 1811 release test

Figure 5  Composition limits and skin-contact release limits are separate requirements, measured on different parts.

One practical note first, since buyers are often handed a certificate rather than a report: a composition result is only as good as what it was run on. A usable report names the component tested, the method, the date, and the batch or lot it represents. A single report covering "the packaging" without identifying the part tells you very little, and a report on an undecorated container tells you nothing about the decorated one.

Two clarifications also matter, because both are commonly got wrong.

Nickel is not in this clause. Nickel is restricted under a different instrument: entry 27 of Annex XVII to REACH, which limits nickel release from articles intended to come into direct and prolonged contact with the skin to 0.5 µg/cm² per week, measured to EN 1811. For a roll-on applicator with a metal ball, that is the relevant test — and it is a release test on a finished article, not a composition limit on packaging material. Confusing the two produces the wrong test on the wrong part.

The 100 mg/kg limit is not new in 2026. Article 11 of Directive 94/62/EC set the same sum-of-four limit, phased to 100 ppm by weight five years after that directive took effect — so it has applied for roughly two decades. PPWR carries it forward and adds derogations, including time-limited ones for recycled glass and certain plastic crates and pallets in closed loops. A supplier presenting this as a new 2026 obligation is telling you something about their compliance literacy.

Results

7
SKUs in the program
5
packaging formats
1.5M+
units on first orders
~80×
spread between smallest and largest SKU

Figure 6  The program as delivered across its first seven months.

The program now covers seven SKUs across five packaging formats. More than 1.5 million units were delivered against first orders, placed in six purchase events over roughly seven months. One format has since been reordered, with repeat volumes expected on two others.

Order sizes ran from roughly 12,000 units on the smallest SKU to over 900,000 on the largest, and the two extremes were placed about a week apart. For a buyer that is the practical test of a consolidated program: whether a small replenishment run still gets scheduled properly while a high-volume line is in production.

What the customer replaced was five separate qualification cycles. One program manager owns the account, one quality standard applies across glass, injection and blow-molded parts and bought components, and one document set comes back regardless of which workshop produced the part.

Buyer Takeaways: What to Build Into a Packaging Brief

Independent of this project, the category practice worth adopting:

  • Specify at component level. Bottle, closure, applicator, ink and coating are separate material declarations. A brief naming only the container leaves the likeliest failure source unspecified.
  • Match the test to the regime. A composition limit and a skin-contact release limit are different tests on different parts. Ask which is run, and on what.
  • Do not accept a migration or safety result as a composition result. They answer different questions.
  • Require change control on decorative materials. An ink or coating swap that looks identical can move a compliance result.
  • Agree the escalation path before you need it. Who is called, how fast, and who holds the cost — settled at award, not during an incident.
  • Test before filling. The cost gap between a non-conformance caught at incoming inspection and one caught after filling is the whole argument for a pre-production gate.
The takeaway

Consolidate by proving one format at a time. The sequencing is the risk control — and what you are buying is a single point of accountability, not floor space.

Frequently Asked Questions

What are the four restricted heavy metals in packaging, and what is the limit?

Lead, cadmium, mercury and hexavalent chromium. The limit applies to the sum of their concentrations in packaging or packaging components, and that sum must not exceed 100 mg/kg. It is not a per-metal limit — the most common misreading. Three metals at 40 mg/kg each would breach it, though none comes close to 100 alone.

Is nickel covered by the 100 mg/kg packaging limit?

No. Nickel is not one of the four metals in that clause. It is restricted separately, under entry 27 of Annex XVII to REACH, which caps nickel release from articles in direct and prolonged contact with skin at 0.5 µg/cm² per week, tested to EN 1811. A roll-on ball or a metal closure detail may well need that test — but it is a different test, on a different part, against a different kind of limit.

Is the 100 mg/kg limit new under PPWR?

No. Directive 94/62/EC set the same sum-of-four limit, reaching 100 ppm by weight five years after it took effect. Regulation (EU) 2025/40 carries the requirement forward from 12 August 2026 and adds derogations, including time-limited ones for recycled glass and for certain plastic crates and pallets operating in closed loops.

How should packaging component testing be planned?

Per component and per material, not per container. Container body, closure, applicator mechanism, ink and coating are separate declarations, and a decorated container needs its decoration covered explicitly. Each component is then mapped to the regimes applying in the destination markets, because composition limits and skin-contact release limits are not interchangeable.

Can a product safety or migration test substitute for a composition limit test?

No. A migration or safety assessment asks what reaches the user or the formula; a composition limit asks what is present in the material. A favourable result on one says nothing definitive about the other, and offering one in place of the other is a common reason documentation is rejected.

結論

Multi-format consolidation works when it is sequenced — one format proven, then scope widened — and when one accountable layer sits above however many manufacturing routes the formats need. It is also a relationship that will eventually be tested by something going wrong. This customer kept building after that happened.

Running several packaging formats across separate suppliers?

We can review the program: component-level specifications, the testing plan for each regime that applies to your markets, and samples against your existing packs.

Talk to our team

Sources

  1. Regulation (EU) 2025/40 on packaging and packaging waste, Article 5 — EUR-Lex.
  2. New EU rules on packaging enter into application — European Commission, 11 August 2026.
  3. Directive 94/62/EC on packaging and packaging waste, Article 11 — EUR-Lex.

    著者について

    As Lead Technical Engineer, Alexandre provides the crucial technical validation for every custom project. He ensures each unique design is structurally sound, functional, and perfectly optimized for manufacturing.

    目次

    最新の投稿