Cosmetic product sampling is the stage of development where the manufacturer produces a small-scale version of your formula, you evaluate it, and you iterate until the product is ready for full production.
It sounds simple. It is not. Sampling is where most first-time cosmetic founders make the most expensive mistakes of the entire launch.
They approve samples too fast because they are excited, they give vague feedback the lab cannot act on, and they miss issues that only appear at scale.
Most sampling guides online treat this stage like a formality. It is the single stage of development where you either catch problems cheaply, or ship them into a production batch that costs far more to fix.
After 30 years in the hair and beauty sector, most recently in private label cosmetics, I can tell you: the brands that launch well are the ones that take sampling seriously.
The sampling process from the founder’s perspective: how many samples to ask for, how to evaluate them, how to give useful feedback, and what to do when a sample does not hit the mark.
Sampling is the iterative loop between the manufacturer’s lab and your approval.
The manufacturer produces a small quantity (typically 50 ml to 500 ml) of the proposed formula.
You receive the sample, test it, document your feedback, and send it back.
The lab adjusts the formula based on your feedback and produces a new sample.
The loop repeats until the sample is approved or the brief is revisited.
Most products go through 2 to 4 sampling rounds before approval. Some go through more if the brief was vague or the customer avatar changed during development.
A common misconception is that more sampling rounds mean a better product. The opposite is usually true.
For the manufacturer, each sampling round is a real cost in lab time, ingredients, and scheduling. Manufacturers aim to get the formula as close to right as possible on the first round, not to iterate ten times. The serious ones see multiple rounds as a signal that something went wrong upstream, not as a normal part of the process.
For the founder, more rounds usually mean the project was underprepared: an unclear brief, an undefined customer avatar, a brand positioning that was never locked. The lab is then asked to fill in the gaps the founder should have filled in before asking for a sample.
The fix for this sits in the stage before sampling.
A founder who has done the strategic foundation work (clear customer avatar, defined brand positioning, specific claims, price tier, distribution channel, reference products) gives the lab enough information to produce a first sample that is already close to the target.
In my own work with the hybrid approach, I study the project’s needs in depth before proposing anything to the manufacturer, even when the client is impatient to see a sample. That preparation often lets me identify an existing formula that already matches the client’s brief closely enough that the first sample comes back approved on the first round. Not always. But more often than most founders think possible.
The rule: do the foundation work first. Then sampling is short and simple instead of long and frustrating.
What you actually receive as a sample depends heavily on which production path you are on.
White label. You receive the finished product as it already exists in the manufacturer’s catalog, usually in the final packaging but with anonymous or neutral labeling. There is no formula adjustment to request. You either approve it as it is, or you pick a different catalog formula. Sample cost is typically the cost of the finished product itself, because it is already a production item.
Hybrid approach. The lab starts from an existing tested formula and modifies specific parameters (fragrance, texture, selected ingredients, active concentrations). Samples arrive in plain lab packaging, usually a simple white jar or bottle. The base formula already has stability data, which shortens the timeline significantly compared to full custom.
Full custom formulation. The lab builds the formula from scratch against your brief. Timelines are meaningfully longer because every ingredient combination is being tested for the first time. Samples also arrive in basic lab packaging.
One important operational detail with full custom: manufacturers typically do not run full stability testing on every sample. If they did, each iteration would take 3 months minimum and the process would stretch for a year or more. Stability testing is usually run only once the formula is nearly locked, on the pre-production sample. This is standard practice and necessary for the process to be practical, but it means the pre-production sample is the one that has to be right before stability work begins.
Each path has its own sampling rhythm. White label is fast and binary (yes or no). Hybrid approach is the most common scenario for first-time brands and moves at a reasonable pace. Full custom is the slowest but gives the most control.
A typical cosmetic sampling process moves through four distinct phases. Each one has a different purpose.
Phase 1: Concept sample. The first prototype, usually based on the manufacturer’s closest existing formula to your brief. This is the starting point, not the goal. Expect it to be 60 to 70 percent right.
Phase 2: Refined sample. The lab incorporates your feedback from phase 1. This should feel noticeably closer to what you described in the brief. Expect 80 to 90 percent right.
Phase 3: Pre-production sample. The formula is locked at this stage. What you approve here is what goes into production. This phase is also where stability testing typically begins.
Phase 4: Production-scale sample. A sample drawn from the first production batch, used to verify that the formula scaled correctly from lab to manufacturing. This catches issues that only appear at volume (viscosity shifts, color changes, fragrance intensity changes).
Skipping phase 4 is one of the most common and most expensive mistakes founders make.
The formula that worked perfectly at 500 ml does not always behave the same at 5,000 liters.
The goal of sampling is to catch every problem before you commit to production, not to fall in love with a sample.
For how sampling fits into the broader development timeline, see the cosmetic product development guide.
This is where most of the expensive mistakes happen. A founder receives a sample, tests it on themselves in the bathroom mirror, and emails the manufacturer "I love it, go to production."
Three months later, the production batch arrives.
The customer feedback reveals issues the founder never caught.
The fix is a structured evaluation process, repeated on every sample.
In most cases, you should receive 1 sample per round, not multiple variants.
Manufacturers do not typically produce 3 different versions of the same formula at once. They produce what they consider the best version based on your brief, and then iterate if you need changes.
Asking for 2 to 3 variants in the same round only makes sense in specific cases: when you are genuinely uncertain between two fragrance directions, when you want to compare two active concentrations, or when the manufacturer themselves suggests it because they see two valid formulation paths for your brief.
Avoid asking for 6 or 7 variants.
Too many samples become impossible to evaluate properly, and the feedback you give becomes muddled. It also signals to the manufacturer that the brief was not clear, which rarely leads to good results downstream.
After the first round, almost always 1 refined sample per iteration. The goal is to narrow toward the final formula, not to keep expanding options.
Evaluate every sample using the same structured framework. A checklist, not a feeling.
The framework I use with founders, distilled from the launches I have guided:
| Criterion | What to check | What to document |
|---|---|---|
| Appearance | Color, clarity, texture on sight | Matches brief? Any unexpected color or separation? |
| Fragrance | Smell in the bottle, on application, after drying | Matches brief? Too strong, too weak, wrong character? |
| Texture on application | How it feels going on, spreadability, absorption | Matches brief? Too thick, too thin, too tacky, too slippery? |
| Performance | Does it deliver the claimed result? | Tested over at least 2 weeks of daily use on real skin or hair |
| Stability at a glance | Has the sample changed since it arrived? | Color shift, separation, fragrance loss over a week on your counter |
| Packaging compatibility | How does it interact with the intended packaging? | Any leaching, fragrance changes, color changes when stored in final packaging |
Fill this out in writing, for every sample, every round. Do not trust your memory, and do not trust impressions.
This is where most advice gets it wrong, in both directions.
The common advice is "do not trust yourself, test with 10 people." That is partially right, partially wrong.
The launches I have guided point to a more specific answer.
For the first few products of a line, the founder’s judgment matters most.
The brand is a reflection of your vision. The first 2 to 5 products are what define the line’s identity.
At that stage, the product has to match the founder’s sense of what the brand should feel like, smell like, perform like. If the founder does not love the hero product, no amount of panel testing will make up for it.
As the line expands, the founder’s taste alone becomes a risk.
Once you are launching product 6 or 7, the founder’s personal preferences start to narrow the line. Maybe the founder loves lavender fragrance and starts putting lavender into every product.
Or the founder likes a specific texture that does not appeal to most of the customer base.
At that stage, you need external feedback to keep the line aligned with the market, not just with your preferences.
The right test panel is small, relevant, and neutral.
Founders often swing too far in the other direction: 10, 15, 20 testers all giving different feedback, creating more confusion than clarity.
Three characteristics make a test panel useful:
The last one is underrated.
Professional testers (hairdressers, estheticians, makeup artists) who use a specific competitor brand daily often bring a bias against any new product. They compare to what they know, not to what the formula actually delivers.
An experienced professional will compare a new formula to a L’Oréal, Wella, Redken, or Davines product they use every day, and conclude it is "not as good," even when the new formula is objectively better on ingredient quality, modernity, or performance. The brand bias is real.
For testing, look for people who can evaluate the product neutrally: professionals who regularly test new products across multiple brands, customers who do not have strong loyalty to a specific competitor, or experienced users with open preferences.
Vague feedback is the single biggest waste of time in the sampling process.
"It feels weird." "I want it to be more luxurious." "The fragrance is off."
None of these tell the lab anything actionable. The lab will guess, produce another sample, and probably guess wrong.
Good feedback is specific, measurable, and written down.
For every issue you want corrected, structure feedback in three parts:
What you observe. The concrete, measurable fact. "The cream feels greasy on my skin even 20 minutes after application."
What you expected. The target, based on the brief. "The brief specified a dry-touch finish within 5 minutes."
What you want changed. A direction for the lab, not a solution. "Reduce the occlusive feel. I do not know which ingredient to adjust, but the current finish is too heavy for the target customer."
Notice the pattern. You describe the problem concretely, point to the brief, and tell the lab what outcome you want.
Working out which ingredient to adjust is the lab’s job.
Feedback is most useful when it describes the problem and the target, and leaves the solution to the people who understand the chemistry.
That principle handles most cases. But there are a few specific patterns that trip founders up even when the intent is right. Worth flagging them before they cost you a sampling round.
Trap 1: Contradicting the original brief mid-process. If your brief said "rich, nourishing cream," do not tell the lab at round 2 that you want it to feel "light and fresh."
Pick a direction and stick with it. Changing direction mid-sampling resets the clock.
Trap 2: Layering personal preferences onto brand decisions. "I personally do not like this fragrance" is different from "this fragrance does not match the brief for the customer avatar."
Be clear which one you are giving. The first is your opinion. The second is a real problem.
Trap 3: Vague approval. Saying "this is close, let us go to production" when you still have doubts is expensive.
If you have doubts, request one more round. One sample round costs a few hundred euros, up to a couple of thousand for full custom work. A bad production run costs tens of thousands.
Trap 4: Approving via email on a Friday afternoon. Sounds silly, but it happens.
Founders approve samples impulsively when they want to hit a launch date. The rushed approval leads to production problems that delay the launch far more than a proper evaluation would have.
For how to prepare the right questions for your manufacturer throughout the sampling process, see the questions to ask a cosmetics manufacturer guide.
Sometimes the first sample is completely off. It does not match what you wanted at all.
This is uncomfortable, especially when the manufacturer has put real work into it.
But accepting a sample that does not fit is worse than having an honest conversation.
Three scenarios, three different responses.
Scenario 1: The sample misses the brief significantly. The brief was probably unclear. Revisit the brief with the manufacturer, clarify what "close to my vision" actually means in concrete terms, and ask for a new round.
Scenario 2: The sample is technically good but commercially wrong. The formula is fine, but your customer will not pay for it. That is a strategic misalignment, not a lab problem. You may need to rethink the product concept before pushing the lab to iterate.
Scenario 3: The manufacturer cannot deliver what you want. Sometimes a lab’s capabilities do not match your specific vision.
If after 3 rounds the samples are still not close, consider whether this manufacturer is the right partner for this product.
For guidance on manufacturer selection and when to consider switching, see the manufacturer selection guide.
Sampling is not the exciting part of cosmetic development, but it is the decisive one.
Brands that sample properly ship products that work in the market.
Rush through it and you ship products that create customer complaints, returns, and expensive reformulation six months after launch.
A failed sample is information about the brief, the formula, or the manufacturer fit, and not a failed project. Treat it as a signal, fix the gap, and only sign off when there are no remaining doubts.
The cheapest time to fix a cosmetic product is before it enters production. Every problem you catch in sampling saves you the far greater cost of catching it after launch.
Treat sampling as the last safety check before committing real money. Get it right, and the rest of the launch becomes much easier.
How long does the sampling process take?
Plan 2 to 4 months from the first sample request to approval of the final production-ready formula. Each sampling round typically takes 3 to 6 weeks: 1 to 2 weeks for the lab to produce the sample, 1 to 2 weeks for shipping, and 1 to 2 weeks for your evaluation and feedback. Skipping any of these compresses quality. Starting stability testing in parallel with the final sampling rounds can save 4 to 6 weeks of total launch time.
How much do cosmetic samples cost?
For white label and hybrid approach, the first 1 to 2 sample rounds are usually included in the project setup fee. Additional rounds typically cost 200 to 800 EUR/USD per sample, depending on complexity. For full custom formulation, each sample round costs 500 to 2,000 EUR/USD because the lab work is more intensive. Some manufacturers offer sample fees as a credit against the first production order. Clarify the cost structure in writing before the project starts. (All cost figures in this article are indicative estimates that vary by manufacturer, region, and project scope.)
Can I ask for multiple variations in one sample round?
Usually no. The manufacturer produces what they consider the best version based on your brief, not 3 versions for you to pick from. Multiple variants in the same round only make sense in specific cases, like when you are genuinely uncertain between two fragrance directions or two active concentrations, and the manufacturer agrees the comparison is useful. Requesting 6 or 7 variants almost always signals an unclear brief, and the feedback you give across them becomes unusable for the next round.
What should I do if I love the sample but my test panel does not?
It depends where you are in the brand journey. For the first few core products of a new line, the founder’s judgment usually carries more weight, because those products define the brand identity. If you genuinely love it and the panel disagrees, investigate why they dislike it before changing direction. For products 6 and beyond in an expanding line, the test panel is a more reliable signal because founder bias tends to narrow the line over time. Either way, listen to what the panel actually says: if specific criticism repeats across testers, that is real feedback. If each tester complains about something different, it is noise, not signal.
Can I skip sampling and go straight to production?
Only if you are doing a pure white label launch with an existing catalog formula you have already tested. For any product involving customization (hybrid approach) or full custom formulation, skipping sampling is not a viable option. The production batch will not match what you imagined, the fixes will be expensive, and any unsold inventory is a sunk cost. The cost of 2 to 3 sample rounds is always lower than the cost of a wrong production run.
What information should I send with a sampling request?
Start with a clear concept brief: what the product does, who it serves, and what result it delivers. Add the technical parameters, including texture, active ingredients, and fragrance direction. Include compliance requirements such as certifications, claims to support, and target markets. Finally, share examples of reference products that match the direction, even if they are from competitors. The more specific the brief, the closer the first sample will be to your target, and the fewer iterations you will need.
How many sampling rounds is normal?
Two to four rounds is typical. One round is either exceptional luck or a too-generous evaluation. Five or more rounds usually signals either an unclear brief, a manufacturer-brand mismatch, or a founder who keeps changing direction. If you find yourself on round four without a clear sample, stop and review what is actually not working. More iterations alone will not fix a structural misalignment.
Keep reading
Honest assessment of AI in cosmetics from an industry veteran with 30 years of experience. What AI can do, what it cannot, and how to use it smartly in your brand.
How to create a haircare product line that actually sells. From hairdresser-turned-consultant: product selection, professional vs retail, and launch strategy.
How to create a skincare product line that actually sells. Strategic guide to product selection, positioning, and sequencing your launch for founders.
