Short answer: sampling is not one sample — it is a chain of approvals where each sample type answers a different question. Run the chain in order and it takes days; skip a stage and it quietly adds weeks to your launch.
Most buyers treat "send me samples" as a single step. In a working OEM project it is a sequence of distinct approvals, and the sample that finally matters — the golden sample — is the one both sides agree is the reference for every production batch that follows. Understanding the chain is the difference between a sample round that converges in two weeks and one that drifts for months.
The five sample types, and what each one is for
| Sample type | What it answers | Who approves |
|---|---|---|
| Stock reference sample | Do I like this direction of formula at all? | Brand |
| Lab sample (custom) | Does the bespoke formula match the brief? | Brand + formulator |
| Revision sample | Did the changes fix what the last round missed? | Brand |
| Golden sample | Is this the permanent benchmark we both keep? | Brand, formally signed off |
| Pre-production sample | Does the real pack, real line, real batch match the golden sample? | Brand, final gate |
The failure pattern is almost always the same: a brand approves a golden sample informally ("looks good, ship it"), then disputes a production batch six months later with nothing in writing to compare against. The golden sample exists precisely to prevent that argument — the manufacturer retains it, and every subsequent batch is compared side by side against it before shipment.
The stage-by-stage chain, and how long each stage takes
A realistic sampling chain for a custom formula looks like this:
- RFQ and stock references (day 0–1). You send the brief; the manufacturer responds with a quote within 24 hours on working days and, if stock formulas fit your brief, reference samples of the closest bases.
- Bench work on the custom formula (day 1–5). The lab builds the formula to your brief: texture targets, active levels, colour, fragrance direction, pH window.
- First lab samples out (around day 7). Under a standard programme, lab samples leave within 7 days of brief confirmation, including two rounds of free revision on the same base. Full custom development from a blank brief runs longer — typically 25–40 days to first samples — because the base itself has to be designed, not just tinted.
- Your evaluation round (your clock). This is the stage that runs longest in practice, and it is entirely yours. Texture on skin, wear through a day, fragrance at different temperatures, compatibility with the pack you chose.
- Revision round(s). Feedback comes back as concrete changes — "viscosity up about 15%, fragrance reduced, colour half a shade warmer" — not "make it better". Vague feedback produces vague samples.
- Golden sample sign-off. Both sides confirm the sample in writing, it is labelled and retained, and it becomes the shipping standard.
- Pre-production confirmation. The first units off the real filling line in the real component are compared against the golden sample. Only after this passes does mass production ramp.
Stages 1–3 sit inside the factory's 7-day sample window. Stages 4–6 are where projects breathe — a brand that evaluates in 3 days converges in two rounds; a brand that evaluates in 3 weeks loses a month per round.
What to evaluate when a sample arrives
A sample evaluated casually always passes; the problems surface later, in the field. A structured evaluation takes an hour and covers:
- Texture straight from the pack — how it dispenses, spreads, and absorbs, at room temperature and after a night in the fridge.
- Wear over hours — not the first five minutes. Foundations oxidise, sunscreens sting eyes late in the day, serums pill under other products.
- Fragrance at two temperatures — a scent balanced at 20°C can turn loud at 35°C, which is exactly where a courier leaves it.
- Colour against your target — daylight, not office lighting, and ideally against the physical reference you sent.
- The pack interaction — does the product separate in the pump, stain the tube, or sit oddly in the jar?
Write down what you find in each round. The written record is what makes revision requests actionable and what makes the final golden sample sign-off meaningful.
Revisions: what "two rounds free" actually covers
Two rounds of free revision is standard on a stock base or a derived formula, and it covers iteration within the same concept — adjusting viscosity, fragrance load, colour shade, active percentage within a workable window.
It does not cover changing the concept mid-stream. If round two reveals you actually wanted a gel-cream instead of an emulsion, or a different preservative system for a market with different rules, that is new development work, not a revision. This is not the manufacturer being difficult; the two cases consume genuinely different amounts of lab time.
The way to stay inside the free rounds is to lock the concept before sampling starts, which brings the conversation back to the brief. A precise brief is what makes the RFQ you send produce a useful first sample instead of a guess.
Common mistakes that stretch sampling into months
- Evaluating samples one person at a time. Collect the whole team's notes, then send one consolidated revision request. Serial feedback turns two rounds into six.
- Skipping the pack question until production. A formula sampled in a lab beaker and approved on skin can behave differently through an airless pump. Evaluate the sample in the component, or accept a late-stage surprise.
- Changing the brief after round one. Every concept change restarts bench work. Lock the brief first — this is one of the quiet time sinks in any manufacturing timeline.
- Golden sample sign-off by message thread. "Approved 👍" in a chat is not a sign-off. A dated confirmation naming the sample code, the pack, and the formula version is what holds up when a later batch is questioned.
- No reverse-engineering discipline. If your sample came from a competitor teardown, make sure the route you chose documents ownership properly — reverse development under NDA produces client-owned formulas; informal copies do not.
FAQ
Q: How long does the cosmetics sampling process take? A: Lab samples for a stock or derived formula typically leave within 7 days of brief confirmation, including two rounds of free revision. Full custom development from a blank brief runs about 25–40 days to first samples. The evaluation rounds in between run on the brand's clock, which is usually the biggest variable.
Q: What is a golden sample in cosmetics manufacturing? A: It is the sample both parties formally approve and retain as the permanent benchmark. Every production batch is compared against it side by side before shipment, so the golden sample — not memory — settles any later dispute about colour, texture or fragrance.
Q: How many sample revisions do I get? A: Standard programmes include two rounds of free revision on the same base — adjustments like viscosity, fragrance load, shade and active percentage. Changing the concept itself (a different texture class or preservative system) is new development work rather than a revision.
Q: Should I approve samples in the actual packaging? A: Yes whenever possible. A formula can behave differently through a pump, a jar or a tube than it does from a lab container, and pack compatibility is much cheaper to discover at the sampling stage than after the filling line is booked.