Ask ten brands where their first launch went wrong & most will point at the sample stage rather than at the factory. The sample is the one moment when a product is still cheap to change, & it is also the moment when buyers make decisions on the least information they will ever have about the product. A texture judged from a photograph, a colour judged on a phone screen, a fill weight approved without a tolerance: each of those becomes a frozen commitment the moment it is signed.
This article sets out how to run a sample round so that the approval holds. It covers what each sample stage can actually prove, how to record sensory impressions in a way that survives translation, which attributes are expensive to change after sign-off, what the approval record has to contain, & how to keep the approved reference usable six months later when the second order is placed. Read it before the first sample arrives & the approval becomes a procedure rather than a rush of enthusiasm.
Why sample approval decides the launch date
A launch calendar is built backwards from a date when stock has to be in a warehouse. That date is usually set by something outside the brand's control: a marketplace listing window, a paid campaign already booked, an influencer seeding schedule, or a trade show. Working backwards from it produces the production run, the artwork approval, the deposit & finally the sample approval, which is where the whole chain actually starts.
Each sample revision round costs more than the days it consumes, because the deposit cannot be released against an unapproved sample & the slot cannot be held without the deposit. A brand that runs three revision rounds instead of one does not simply lose two weeks; it enters the production queue two weeks later, & queues are not linear.
The cash consequence is worth stating plainly for a first-time seller. Money is already committed to the campaign, the creative & often the packaging tooling before a single unit exists, & none of it produces revenue while a sample sits unapproved on a desk.
What a sample is actually for at each stage
Sampling is not one event. It is a sequence of three or four artefacts, each of which answers a different question, & confusing them is the single most common cause of a wasted round. A laboratory sample answers whether the formula can exist. A scale-up or pilot batch answers whether it can exist at the batch size you intend to buy. A stability sample answers whether it will still be within specification after the labelled shelf life. A production sample lifted from the first commercial run answers whether the line reproduces what the lab made.
Brands routinely approve a laboratory sample as though it were a pilot sample. The lab version is often mixed by hand, filled by hand & matured for days rather than weeks, so its texture can be smoother & its colour more even than anything a production line will produce. When the commercial batch arrives looking slightly grained, both sides argue from different evidence, because nobody agreed in advance which stage the approval covered.
The practical rule is that a signed approval should name the stage it applies to. Approving a lab sample is a legitimate decision, provided it is written as one: the brand accepts the formula direction, with a second approval required against the pilot batch before the run is released. That single sentence prevents most disputes in this area.
Reading a sample against the brief rather than the photo
The brief is the only document a sample can be judged against, & it usually contains fewer numbers than it should. "Light, fast-absorbing, non-greasy" describes a direction, not an acceptance criterion. When a sample arrives, the brand has nothing to measure it with except a feeling, & feelings are difficult to argue about politely.
The remedy is to convert the brief into attributes with a scale before sampling begins. Absorption time in seconds, surface tack on a five-point scale, gloss level, spread rate, drag, slip after dry-down. None of this requires a laboratory, only that the descriptors are fixed in advance & the same person evaluates each round.
Photographs create a separate problem. A phone camera renders colour differently from the eye, adjusts white balance without asking, & flatters gloss levels depending on the light in the room. A sample photographed under a domestic bulb & viewed on a calibrated monitor two continents away is not evidence. Where a photograph has to carry a decision, it needs a grey reference card in frame, indirect daylight, & a note of the device used, so that the next round can be shot the same way.
One more comparison to make explicitly is against the reference product. If the brief named a product already on the market, the sample should be assessed side by side against it rather than from memory. Brands that skip this step often discover on the third round that they have been describing the reference product inaccurately from the start.
The sensory sheet: recording what you feel
Write the sensory sheet before the sample arrives, not after. It should have one line per attribute, a scale with words anchored at each end, a column for the reference product, & a column for the sample. It takes an afternoon to build & it is the difference between a decision & an opinion.
Keep the vocabulary narrow & repeatable. Words that survive translation best are physical rather than emotional: thick, thin, wet, dry, tacky, smooth, grainy, cool, warm, glossy, matte, tight, slippery. Words that do not survive translation include luxurious, elegant, premium, silky & spa-like, all of which mean something different to every reader.
Severity matters as much as direction. A grainy texture that disappears after thirty seconds is a note; a grainy texture that persists is a rejection. The sheet should therefore record the point in the routine at which an attribute is observed: on application, at one minute, at five minutes, & after the product has sat for an hour.
Where the product carries a fragrance, record the intensity on the same scale, & record it on skin rather than on paper. Fragrance behaves differently on a warm surface, & the load that reads as light in a vial can read as heavy once applied. This is the point at which the allergen position becomes relevant: the twenty-six substances that must be named on a label above their applicable threshold are declared by the fragrance house, & a brand that changes the load after sampling may need a fresh declaration.
Colour, odour, fill weight: the drift subjects
Three attributes deserve separate attention because they change after approval & are expensive to correct later. The first is colour. A tinted capsule shell, a coloured gel or a tinted emulsion will look different under different lighting, & some colourants fade when exposed to light over a shelf life. Approving colour from a single photograph at a single moment is the weakest form of approval there is.
The second is odour. A base raw material can carry its own note, particularly botanical extracts & oils, & that note varies between lots. A brand that approves an odour profile without asking what the natural variation looks like will receive a batch that is inside the raw material specification & outside the brand's expectation.
The third is fill weight. A claim printed on a pack is a promise about quantity, & the tolerance around the target is a manufacturing parameter rather than a marketing choice. Approve the target & the tolerance together, & check whether the tolerance is stated as a percentage or an absolute weight, because a two per cent band on a small fill is a very different number from a two per cent band on a large one.
Where a volatile fragrance or a high alcohol content is involved, the approval round is also the moment to settle the transport classification. Flash point decides whether the finished goods travel as a flammable liquid, which changes the carrier, the documentation & sometimes the pack format. Discovering that at the booking stage adds cost that no sampling decision can recover.
When to approve, when to ask for a revision, when to stop
Approval is the right answer when the sample meets the acceptance criteria in the brief, even if it is not the brand owner's personal favourite. The purpose of a brief is to remove taste from the decision. A sample that satisfies every stated criterion & simply feels unexpected is a brief problem, not a factory problem.
A revision is the right answer when one attribute falls outside the agreed band & the cause is identifiable. Revisions work when they are narrow: change the fill weight, change the fragrance load, reduce the drag. Revisions fail when they are broad, because a factory given eight simultaneous changes will adjust the process in ways that alter attributes nobody asked about.
Stopping is the right answer when the brief itself turns out to be wrong. If two rounds have produced products that satisfy the written criteria & the brand still does not want either, the issue sits in the brief. Continuing to sample is an expensive way to avoid rewriting one page.
Most factories include a defined number of revision rounds in the sampling fee, & the number is worth confirming in writing. It is normal for a first round to be charged & for one or two revisions to be included. It is not normal for a factory to charge a fresh development fee for a minor adjustment to an attribute that was inside the original scope.
The approval record: what the signature must carry
An approval is only useful if it can be read again by someone who was not in the conversation. The record should name the product, the formula version, the sample stage, the sample reference or code, the date of dispatch, the date of evaluation, & the acceptance criteria that were applied. It should state what was approved & what remains open.
The most commonly omitted line is the version. A factory may hold three formula versions of the same product within a month, & an approval that does not name one can be attached to any of them. The same applies to packaging: an artwork approval that does not name a revision number is not an approval, it is a starting position.
A written record also creates the audit trail a brand may need later. If a batch is questioned a year after launch, the approval document is what shows which specification was in force at the time of manufacture, & it is the difference between a storage question & a formulation question.
Keeping the approved reference alive after sign-off
The approval signature is where most brands stop, & the reference is where they should start. An approved reference sample should be retained by both sides, sealed, labelled with the formula version, the sample code & the date, & stored under the conditions the label claims. Without it, the second order is judged from memory against a product that no longer exists as a physical object.
Storage conditions change what a reference can prove. A sample kept on a shelf in an office for eight months can still confirm colour & fill weight, but it cannot be used to argue that the formula has drifted if the storage was warmer & brighter than the label specifies. If the reference is going to carry that weight, it needs the same storage as a retention sample.
A useful routine is to review the reference at the point of each reorder rather than waiting for a complaint. Comparing a new batch against the retained reference at one minute, five minutes & one hour takes fifteen minutes & catches drift while the batch is still in the factory & still fixable.
Sample stages: what each one can settle
The table below sets out the usual sampling sequence for a cosmetic project, what each stage can reasonably settle, & the decision that should not be taken on that stage alone.
| Stage | What it proves | What it cannot settle | Typical owner |
|---|---|---|---|
| Reference comparison | That both sides describe the same target product | Whether the target can be made at your batch size | Brand |
| Laboratory sample | That the formula exists as a direction | Texture & colour at production scale | Factory |
| Revised lab sample | That a named attribute can be moved | Whether other attributes stayed unchanged | Both sides |
| Pilot or scale-up batch | That the product survives the real process | Whether it holds for the full shelf life | Factory |
| Stability sample | Behaviour under accelerated & real-time conditions | Anything about sensory preference | Factory |
| Pack compatibility sample | That the pack & the formula coexist | Filling accuracy on the line | Both sides |
| First production sample | That the line reproduces the approved product | Nothing further at this stage | Factory |
Those stages are a working sequence rather than a fixed rule, & a proven base formula can skip one or two of them while a full development project may add a round. What does not change is the need to name the stage on the approval, because the evidence behind each row is different.
What to put in the brief
Sampling is decided long before the sample arrives, in the wording of the brief. State the acceptance criteria, the number of revision rounds expected, the reference product with the attributes you are copying & the ones you are changing, & who inside the brand holds the approval right. If two people have to agree internally, say so, & put their sequence in the brief rather than discovering it during a round.
- The formula base or reference product, named precisely enough to be sourced.
- The attributes being matched & the attributes being deliberately changed.
- Sensory targets stated on a scale, with the evaluation point in the routine.
- The fill weight target & the tolerance, expressed in the units the label will use.
- Colour & odour expectations, including the variation you will accept between lots.
- Whether a fragrance is present, & whether an allergen declaration has to be issued with the sample.
- The number of revision rounds expected & what counts as a new development.
- The named approver, the deputy, & the internal sign-off sequence.
Then decide in advance what happens if the sample is late. A factory that misses a sampling date should be asked whether the delay is a capacity problem or a materials problem, because the two have different fixes & only one of them can be solved by paying more.
Where our capability fits
We sample in named stages & we say which stage each sample is, so that an approval covers the evidence actually available at the time. A laboratory sample is issued with its formula version, its fill weight & the raw material lots used. A pilot batch is issued with the process parameters as run, & a stability sample is issued with the conditions & the elapsed interval rather than a summary line. Where a fragrance is in the formula, the allergen declaration & the IFRA documentation travel with the sample rather than arriving with the commercial invoice.
We work with brands on private label (your brand on a proven base), OEM/ODM contract manufacturing & full custom formula development, & the sampling record is the same for all three. Our lines include custom softgel capsule face serums, a hyaluronic acid capsule serum & a fullerene capsule range, where shell colour, fill weight & shell integrity are all approved at a named stage.
Tell us the attributes you will judge a sample on & the date your stock has to be in the warehouse, & we will come back with the sampling sequence that fits the calendar rather than the other way round. Start at our inquiry page, email capsulebiotech@163.com, or message the right team on WhatsApp: Capsule Series (capsules only) +86 158-7655-5319, Cosmetics Series (serum, hair care, body care, fragrance) +86 158-7655-5319. How a brief becomes a gold sample is set out in our sampling process guide, & how brands test before committing to a full run in our article on pre-run testing.
General manufacturing information, not regulatory or legal advice. Sampling norms, labelling duties, allergen thresholds & transport classifications vary by market & by product, & the applicable requirements should be confirmed for each destination before a product is placed on sale.