Formulation Sample Feedback: How to Write It and When to Lock the Formula

AI-generated image.

The short answer: comparing means matching conditions, and confirming means knowing what cannot be undone

There are two jobs in a sample round. One is to record what you observed under identical conditions in language specific enough to become the brief for the next sample. The other is to close, on time, the decisions that get more expensive the longer they are deferred. The first job looks like a matter of taste but is really a matter of record-keeping. The second looks like a matter of scheduling but is really a matter of law and paperwork.

Start with the second. A functional cosmetic must, under Article 4(1) of the Cosmetics Act, be reviewed by the Minister of Food and Drug Safety for safety and efficacy on a product-by-product basis, or be the subject of a submitted report. The same paragraph adds a second sentence: "The same shall apply where any submitted report or reviewed matter is modified." In other words, changing the formula after the review or report is complete means going through the procedure again. Changing the fragrance can change the ingredient declaration. Changing the container reopens compatibility with the contents. If a late round shakes the formula because the texture finally feels right, it also shakes testing that is already finished and labels that are already printed.

This article covers how to compare samples under identical conditions, how to turn a vague impression into a development request, and exactly which provision each costly decision is attached to: functional claims, fragrance, container and shelf life.

The scope is the law of the Republic of Korea. The Act, Enforcement Rule and notices cited here were checked against the current text on the Korean Law Information Center on 17 October 2026. This is not legal advice for a specific matter, and it makes no claim about EVAS commercial terms, facilities or testing records.

Key points

• Comparison means matching conditions. If the amount used, the area, the involvement of water or the observation time differ, you compared two situations rather than two samples.
• Change one or two things per round. Moving fragrance, viscosity and residual feel at once makes it impossible to tell what worked.
• Sample notes are not evidence of efficacy. For functional review, the notice sets out how human application study data must be conducted and written.
• Changing the fragrance can change the ingredient declaration. Allergenic substances among the constituents of a fragrance cannot be grouped as "fragrance"; their names must be listed.
• Functional products move the lock point earlier. Article 4(1) of the Act requires the same procedure again whenever a reviewed or reported matter is modified.
• The report route that replaces review has conditions. Article 10(1), item 2 of the Enforcement Rule requires the ingredient type, specification and content, the efficacy, the standards and test methods, the dosage and administration, and the dosage form to be all identical.
• Using product made for review is narrowly allowed. Article 8-2 of the Regulation on the Review of Functional Cosmetics requires both compliance with good manufacturing practice and an "appropriate" review result, and caps it at three manufacturing units by lot number.
• Once the formula is fixed, documents follow. Article 11(1), item 2 of the Enforcement Rule requires manufacturers to prepare and keep the manufacturing control standards, product standards, manufacturing records and quality control records.
• Samples cannot be sold to consumers. Article 16(1), item 3 of the Act prohibits selling cosmetics manufactured or imported for consumers to try in advance.
• Tester packaging may carry a reduced set of particulars. Article 19(1), item 2 of the Enforcement Rule and its proviso apply, and there "price" means an indication such as sample or not-for-sale.

1. If the conditions differ, it is not a comparison

When two samples are discussed, the first thing that drifts is not the formula but the conditions. If the amount used, the area of application, the involvement of water or the time of observation differ, there is nothing left to compare.

• Fix the amount. "One pump" and "a pea-sized amount" differ from person to person. Use the same tool to dispense the same quantity where possible.
• Fix the area. The back of the hand, the cheek and the inner forearm differ in sebum and temperature.
• Fix whether water is involved. For cleansers and shampoos, the volume and temperature of water change the impression of the lather considerably.
• Fix the observation time. Residual feel immediately after application, after one minute and after ten minutes are three different pieces of information.
• Note the temperature and season. The same lotion is judged differently on a winter morning and a summer evening.

Rather than concluding from a single impression, record the conditions under which you noticed a difference. These notes do not replace a consumer usability study or efficacy testing. They exist to narrow the development direction.

2. Record the sequence of use in steps

An impression written as a whole gives no indication of what to fix. Breaking the motion into steps turns the same sentence into a request.

• Initial form when dispensed: flow, tack, clumping, color
• Resistance while spreading: initial slip, drag in the middle, the area it covers
• Changes during use: how fast and how fine the lather builds, when emulsification begins, any warming
• Finish: the feel after rinsing or the residue after absorption, gloss, adherence

The same questions do not apply to every product. For a cleanser the critical interval is the moment it meets water; for a shampoo it is lather and rinsing; for a lotion it is absorption and residual feel. Where product type itself changes formulation latitude is covered separately in the article on body wash and body lotion.

3. Turn vague wording into a specific request

"Too heavy" is an impression. "I want less residue at the finishing stage rather than during spreading" is a request. The difference is whether the formulator can identify a variable to move.

• "Good lather" → separate the volume, the bubble size, the resilience and the feel during rinsing.
• "I dislike the fragrance" → say whether it is the opening, what rises during use, or what remains after drying.
• "It feels stiff" → distinguish resistance coming out of the container from resistance on the skin.
• "Moisturizing" → say whether you mean hydration immediately after application or the absence of tightness hours later.

One caution when choosing words: a term used in a sensory note does not automatically become product copy. Statements that go out as labeling or advertising follow separate substantiation requirements, and that structure is set out in the article on advertising substantiation.

4. Change one or two things per round

If fragrance, viscosity, spreadability and residual feel all move at once, the next sample cannot tell you what improved. When there are many things to change, it is faster to rank the one or two that matter most to the concept and discuss those with the formulator.

Three criteria help with the ranking.

1. Does it decide the concept? If sensory experience is the core of the concept, close the sensory question first.
2. Does it drag other items with it? Changing viscosity moves spreadability and residual feel; changing fragrance moves the ingredient declaration.
3. Is it expensive to undo? The functional, container and shelf-life questions below belong here.

How much latitude there is to change also depends on whether you start from an existing base formula or build a new one. That difference is set out in the article comparing base formulas and custom formulation.

5. Where a photo or a short video is needed

Some moments are hard to describe. How the sample dispenses, how it spreads on skin, how lather is built, the instant it emulsifies with water. For these, a short clip is more precise than three paragraphs.

When filming, fix the lighting and background and shoot from the same distance. And images of actual samples must be kept clearly separate from concept images. If a concept image is used as material showing the product's actual appearance or effect, it becomes a labeling and advertising problem.

6. What sample notes cannot substitute for

A record of ten people trying a sample and liking it is useful development material, but it is not evidence of efficacy. In functional cosmetics review, the efficacy data are efficacy test data and human application study data, and Article 5, item 1(c) of the Regulation on the Review of Functional Cosmetics sets out the requirements in detail.

• Efficacy test data must be produced by a domestic or overseas university or specialized research institution and issued by the head of that institution, or be data the government of the country of development received or approved, or data published in a specialist journal indexed in the Science Citation Index.
• Human application study data follow the same requirements, and the conduct standards and writing method apply the Regulation on Substantiation of Cosmetic Labeling and Advertising mutatis mutandis.
• Safety data must be tested in accordance with the Good Laboratory Practice standards, and human patch tests and cumulative human patch tests must be conducted at a domestic or overseas university or specialized research institution, performed and evaluated under the guidance and supervision of a person with at least five years of relevant testing experience.

One terminology note is worth keeping. The "safety data" in a review dossier are toxicity, irritation and sensitization studies. The "stability" discussed when setting a use-by date is about whether the product changes over time. They sound alike and are not the same data.

7. Changing the fragrance changes the ingredient declaration

Fragrance is the item most often changed in the final round, and it connects straight to the ingredient declaration. A fragrance substance may be listed as "fragrance," but where its constituents include an allergenic substance designated by the Minister of Food and Drug Safety, it cannot be grouped that way and the name of that substance must be listed. The list of those substances and the concentration thresholds sit in a notice, and the threshold differs between rinse-off and leave-on products.

Two things go wrong in practice.

• The same fragrance composition can produce different declarations depending on product type. A substance grouped away in a body wash can surface by name in a body lotion.
• Changing the fragrance when the label artwork is about to go to print changes the number of lines in the ingredient list, which means redoing the layout.

The criteria on the ingredient side are in the article on selecting ingredients, and the warning statements that go on the label are in the article on precautions for use.

8. Functional products move the lock point earlier

To sell a product as a functional cosmetic, Article 4(1) of the Cosmetics Act requires a product-by-product review of safety and efficacy by the Minister of Food and Drug Safety, or the submission of a report. The second sentence of that paragraph is what matters in practice: "The same shall apply where any submitted report or reviewed matter is modified."

Where a review is sought, the documents to be submitted are listed in Article 9(1) of the Enforcement Rule.

1. Data on the origin and development history
2. Safety data: single-dose toxicity, primary skin irritation, eye mucous membrane or other mucous membrane irritation, skin sensitization, phototoxicity and photosensitization, and human patch test data
3. Efficacy or functionality data: efficacy test data and human application study data
4. Data substantiating the sun protection factor and the UVA protection grade (only for products with a UV protection function)
5. Data on standards and test methods (including a specimen)

Note that the fifth item says "including a specimen." This is not a paper-only procedure; an actual product is submitted with it, and that specimen has to be made from the confirmed formula.

There is a separate procedure for changes. Article 9(3) of the same Rule requires a person seeking to modify a reviewed matter to submit an application for modified review, attached with the previously issued review result notice (except where it was issued as an electronic document) and documents substantiating the reason for the change.

The way review data must be written also assumes a fixed formula. Article 10(1) of the Regulation on the Review of Functional Cosmetics requires the blending purpose, ingredient name, specification and quantity to be stated for each ingredient, with the quantity stated in principle "per 100 millilitres" or "per 100 grams." Colorants, fragrance substances, emulsifiers, pH adjusters and viscosity adjusters without a designated usage limit may be stated as a suitable quantity. Article 9 of the same notice requires the product name to differ from that of a functional cosmetic already reviewed, and Article 11 requires the dosage form to be stated using the forms set out in the general rules of the Standards and Test Methods for Functional Cosmetics.

In short, for a functional product "confirming the sample" is not agreeing on texture. It is closing the ingredient types, specifications and quantities, the product name and the dosage form at the same time. If you have whitening, wrinkle improvement or UV protection in mind, it is safer to count the sample rounds backwards from the review schedule. Where the numbers on a sunscreen come from is covered in the article on SPF and PA, and a case that turns on whether a claim is functional at all is in the article on hair-loss claims for shampoo.

9. The report route also comes with conditions

Not every functional cosmetic goes through review. Article 10(1) of the Enforcement Rule sets out which products are reported instead of reviewed. The first category is a functional cosmetic whose type and content of the active ingredient, efficacy, dosage and administration, and standards and test methods are the same as a product notified by the Minister of Food and Drug Safety.

The second category is the one used most often in development. It is a product for which all of the following are the same as a functional cosmetic already reviewed:

• the type, specification and content of the ingredient producing the effect (concentration, where liquid)
• the efficacy
• the standards [excluding the standard for pH] and test methods
• the dosage and administration
• the dosage form

This route carries a requirement as to who you are: it is limited to cases where the manufacturer is the same (only where the manufacturer designed, developed and produced the product), where the responsible distributor is the same, or where the research institution that obtained the review is the same. On dosage form, item (e) treats solutions, lotions and creams as the same dosage form for functional cosmetics other than those for tanning and UV protection (Article 2, items 4 and 5 of the Enforcement Rule). So the answer to "does switching a lotion to a cream change the dosage form?" depends on which function is involved.

What this means at the sample stage is plain. Keeping the report route open makes even a "small" adjustment to a content level a non-trivial decision. Standards and test methods have to match apart from pH, so swapping a thickener for texture reasons becomes something to check.

10. Product made for review is capped at three manufacturing units

Whether product manufactured ahead of a review can be used is another common question. Article 8-2 of the Regulation on the Review of Functional Cosmetics sets the conditions. Where cosmetics manufactured for the purpose of a review or modified review meet both of the following, they are deemed to have been manufactured after the review, notwithstanding Article 15, item 1 of the Cosmetics Act:

1. they were manufactured in compliance with the good manufacturing practice standards under Article 11(2) of the Enforcement Rule of the Cosmetics Act; and
2. the review or modified review result was notified as appropriate.

Even then, the quantity may not exceed three manufacturing units counted by lot number. Building a comfortable stock before the result arrives is not a workable plan. What a GMP conformity assessment does and does not prove is covered in the article on CGMP.

11. Container and shelf life run alongside the samples

Sample evaluation is usually done by dispensing from a neutral container, but the real product ships inside its packaging. Choosing the container after the texture is locked therefore shakes two things at once.

• Compatibility between contents and container. Migration of fragrance and color, shrinkage and swelling, and pump output are checked once the formula is set, but it is safer to narrow the container candidates during the sample rounds. When to run that check is covered in the article on packaging compatibility.
• Use-by date and period after opening. These numbers do not come from sensory results; they come from stability evaluation, and a change in formula restarts it. The structure of that evidence is in the article on shelf life and period after opening.

Finished-product testing is a further axis. Whether a texture you liked in a sample passes the safety management standards for distributed cosmetics is checked separately, and that list is in the article on finished-product test reports.

12. Once the formula is confirmed, documents follow

From the moment the sample rounds end, the formula becomes a controlled item. Article 11(1), item 2 of the Enforcement Rule of the Cosmetics Act requires a cosmetics manufacturer to prepare and keep manufacturing control standards, product standards, manufacturing records and quality control records, including in electronic form.

The product name in the product standards also connects to the product name in production performance reporting. The field-by-field basis for those forms is in the article on production and ingredient list reporting. Rights in the formula itself, meaning who may use a developed formula and within what scope, are better settled before samples arrive, and that question is covered in the article on who owns the formula.

13. Samples cannot be sold, and their labels differ

How a brand may handle the samples it receives is also in the provisions.

Selling is prohibited. Article 16(1), item 3 of the Cosmetics Act prohibits selling, or storing or displaying for the purpose of sale, cosmetics manufactured or imported so that consumers may try them in advance for promotional or sales purposes rather than for sale. The proviso to that paragraph limits item 3 to cosmetics sold to consumers. Reselling in-store testers or giveaway product to consumers is not a workable arrangement.

The packaging particulars, on the other hand, may be reduced. The proviso to Article 10(1) of the Act allows packaging prescribed by Ordinance of the Prime Minister to carry only the name of the cosmetic, the trade name of the responsible distributor and the customized cosmetics seller, the price, the lot number and the use-by date or period after opening. Article 19(1), item 2 of the Enforcement Rule designates as such packaging "the packaging of cosmetics manufactured or imported so that consumers may test or use them in advance for the purpose of product selection and not for sale." The proviso to the same paragraph states that, for packaging under item 2, "price" means an indication such as sample or not-for-sale. The full structure of what can be reduced on small packaging and samples is in the article on small packaging labels.

14. A feedback format and a working checklist

Feedback that leads to the next sample needs four fields: what you liked, what you would change, why, and the conditions that must be preserved. Add the record of conditions to that.

• Evaluation conditions: amount, area, whether water was used, observation time, number of evaluators
• Sample identification: formula code and date received, so samples do not get mixed up
• Priority: the one or two items to change this round
• Fixed conditions: what must not change, such as the presence or absence of a particular ingredient

Before sending, check the following.

1. Were the two samples compared under the same conditions, and were those conditions written down?
2. Is it written as a request rather than an impression? Can a variable be identified?
3. Has this round been narrowed to one or two items to change?
4. Is a functional claim planned? If so, is the formula lock point aligned with the review or report schedule?
5. Has it been checked whether a fragrance change affects the ingredient declaration?
6. Have the container candidates been narrowed, and is there a date for the compatibility check?
7. Is the evaluation needed to set the use-by date inside the overall schedule?
8. Have the handling rules for samples and testers (no sale, reduced labeling) been shared with the sales and marketing side?
9. Is one person responsible for consolidating feedback?

What to organize before a first consultation is in the article on writing a product brief, and the stage where quantity and conditions are set is in the article on small-batch production and MOQ.

What this article did not check

• The number of samples provided, the cost, the number of rounds and the schedule are project conditions, not matters set by law. This article does not state them.
• The scope of data exemptions in functional cosmetics review sits in Article 6 of the notice and is outside this article.
• The list of the twenty-five allergenic fragrance substances and the detail of their concentration thresholds are not covered here.
• The design and acceptance criteria of stability evaluation are a separate subject; this article only treats its schedule as connected to the formula lock.
• Registration and notification procedures in other countries are outside this article. If you are planning exports, the importing country's rules must be checked separately.

Frequently asked questions

How many sample rounds should we expect?

The law sets no number. In practice the number of rounds is decided by how many items you want to change and how tightly those items are coupled. Changing one or two per round means more rounds but a clear read; changing several at once means fewer rounds and moving on without knowing the cause.

If we like the texture, can we go straight to production?

Texture is one axis of confirmation. Production conditions are set once the functional question, the ingredient declaration, container compatibility, the use-by date and the finished-product test items are settled together.

Is it fine to adjust only the fragrance at the end?

Fragrance connects to the ingredient declaration. If an allergenic substance designated in the notice exceeds the threshold among the constituents of the fragrance, it cannot be grouped as "fragrance" and the name has to be listed, which can change the label artwork. For a functional cosmetic, the modification procedure for the review or report has to be considered as well.

Can we put the results of ten people trying the sample on the product page?

That record is development material. Copy that goes out as labeling or advertising follows substantiation requirements, and the conduct standards and writing method for human application study data are set by notice. Presenting a sensory survey as evidence of efficacy is not advisable.

Can leftover samples be sold cheaply to consumers?

Article 16(1), item 3 of the Cosmetics Act prohibits it. Cosmetics manufactured or imported so that consumers may try them in advance cannot be sold to consumers, nor stored or displayed for the purpose of sale.

Can we manufacture ahead of a functional cosmetics review?

Conditionally. Article 8-2 of the Regulation on the Review of Functional Cosmetics deems product to have been manufactured after the review only where it was made in compliance with good manufacturing practice standards and the review or modified review result was notified as appropriate, and it caps the quantity at three manufacturing units counted by lot number.

Next steps

Comparing samples is not about narrowing down a preference. It is about closing decisions. Look under the same conditions, change one thing at a time, and close the items that are expensive to undo first; the rounds get shorter and the schedule becomes predictable.

EVAS runs planning, design, research, manufacturing, logistics and brand content together in house. You can look through the formulation library for a development direction, or tell us about the samples you are comparing through a consultation request. Even a note on the conditions and what you want to change is enough for us to help set the criteria for the next round.