Korean ODM Sample Evaluation: 9-Step Protocol (2026)
By the ALTA MEET editorial team | K-beauty ODM consulting
A Korean ODM ships you three glass vials, a 1-page tech sheet, and an email that says "Please confirm by Friday." It is the most expensive Friday of your launch. Most indie K-beauty founders look at the samples, smell them, write back "looks great" or "can we make it lighter," and lose the next eight weeks rebuilding a formula that they could have nailed on the first round if they had run a real evaluation protocol.
Sample evaluation is the gap step. Korean ODMs run rigorous internal QA. Founders are taught to brief and to read a quote. Almost no one teaches you what to do with the physical samples sitting on your desk. This guide is the protocol we use at ALTA MEET when we open a sample box, and the order matters: each step screens out a class of error so the next step is meaningful. (For the cost lines that follow sample approval, see how to read a Korean ODM quote line-by-line.)
What's Actually Inside a Korean ODM Sample Round
Before any evaluation, you need to know what the package on your desk should contain. A complete first-round sample shipment from a Korean ODM typically includes:
Two to three formula candidates in unlabeled glass or PET vials (often 30 ml or 50 ml fill)
A 1-page tech sheet per candidate: INCI list, pH, viscosity at 25°C, color and odor description, base preservative system
The ODM's internal organoleptic and physico-chemical pre-screen data (sometimes a 4-week or 6-week mini-stability table) (For how to read the stability test report itself, see Korean ODM stability testing.)
A cost-per-unit estimate at your target MOQ, often labeled "indicative" because materials cost is still locked to a 30-60 day quote
If any of these is missing, do not begin evaluation. Ask the ODM to send the missing piece. A sample without a tech sheet is uninterpretable; a tech sheet without pH and viscosity numbers is incomplete. Korean ODM labs measure these by default, so a missing number means the lab held it back, not that it does not exist.
A note on quantity: 30 ml is enough for organoleptic, pH, viscosity and a small wearer trial. It is not enough for a full ICH Q1A accelerated stability run. If you intend to perform any third-party stability test, request a 200-300 ml master sample in a separate request before you open the evaluation vials.
Step 1: The Unboxing Audit
Open the box on a clean, flat, dust-free surface. Do this with a notebook open and a phone camera in hand. Photograph:
The outer shipping label, including the country-of-origin stamp and any customs paperwork
The cold-chain indicator, if present (a small color-changing sticker that confirms the box did not exceed 30°C in transit)
The vials in the orientation they shipped in (settling, separation, oil films at the top)
The label or sticker on each vial as received
This step takes four minutes. It catches three issues that quietly contaminate the entire round: a sample that froze in transit (cloudy, separated, lower viscosity than spec), a sample that overheated (degraded actives, off-color, sour odor), and a labeling mismatch where the ODM swapped candidate A and B in the email but not on the vial. Walk into the rest of the protocol with the unboxing photos archived. If you reject the round later, the unboxing photos are how you prove the cause was a shipping or labeling issue, not a formulation one.
Step 2: Organoleptic Evaluation (Appearance, Color, Odor, Texture)
Organoleptic evaluation, sometimes called sensory evaluation, is the structured human inspection of a product's physical attributes. Korean ODMs do this in-house with a trained panel; you are running a single-evaluator version of the same protocol, which is fine for first-round screening as long as you record what you observe and check it again 72 hours later under the same conditions.
Score four attributes on a 1-to-5 scale against your brief:
Appearance: clarity for a watery essence, opacity for a cream, homogeneity for a fluid. Any visible particulates, oil droplets, or color streaks score 1.
Color: hue and intensity. Korean ODMs typically describe target color in CIE Lab coordinates. Ask for the target Lab values and compare to your sample under daylight (around 5,000-6,500K). A delta E (ΔE) of less than 2 is generally acceptable for batch consistency; a ΔE of more than 3 is visible to the naked eye and should be flagged.
Odor: base odor (the raw smell of the carrier and actives) versus added fragrance. Smell the sample at room temperature, then warm a small amount on the back of your hand and smell again. Heat releases volatile fragrance compounds and unmasks any off-notes from oxidized oils or stressed botanicals.
Texture: slip on the back of the hand, pickup from the vial, breakdown on application, finish (tacky, dry, dewy). Korean ODMs often target a specific finish profile derived from the reference product you submitted in the brief.
Record numbers, not adjectives. "Feels too thick" is not actionable for an ODM. "Pickup is 3 of 5 versus target 5 of 5; spreads in 8 seconds versus target 4 seconds" is.
Step 3: pH and Viscosity Verification
This is the first step that requires equipment. You can do it with a calibrated pen pH meter (USD 50-150) and a Brookfield-equivalent viscometer rental (a few hundred dollars for a weekend rental at a contract lab), or you can outsource it to a US contract lab for around USD 80-150 per sample for both measurements together.
Cross-check the sample's measured pH against the ODM tech sheet:
A pH drift greater than 0.3 units between the ODM's at-batch reading and your at-arrival reading is a flag. Possible causes: in-transit temperature stress, incompatible preservative-active interaction, or a measurement error on either side.
Confirm the pH is appropriate for the active payload. Vitamin C as L-ascorbic acid wants 2.5-3.5 to stay reduced. Niacinamide is comfortable from 5.0 to 7.0. Retinol wants 5.5-6.5. AHA chemical exfoliants in the US require labeling guidance from the FDA when pH falls below 3.5 and free acid concentration is high.
For viscosity, accept the ODM's measurement as the reference if your at-arrival reading matches within 10 percent at the same temperature (25°C is the standard). A larger drift is usually emulsion instability, which Step 4 catches.
Document the readings in your evaluation log. These numbers become the baseline you compare future batches against.
Step 4: Stability Validation (ICH Q1A Conditions, Adapted for Cosmetics)
ICH Q1A(R2) is the pharmaceutical industry's stability testing framework. Korean cosmetic ODMs and global labs use it as the reference protocol for cosmetic stability, even though cosmetics are not pharmaceuticals. The two cosmetic stability conditions that matter:
Accelerated: 40°C ± 2°C / 75% RH ± 5% RH for at least 6 months. Six months under this condition is generally accepted as predictive of a 24-month shelf life under normal use, though the relationship is empirical and active-dependent.
Real-time / long-term: 25°C ± 2°C / 60% RH ± 5% RH for at least 12 months, often extended to 24 or 36 months. This is the test the shelf-life claim on your future packaging is anchored to.
For first-round sample evaluation, you are not running the full 6-month protocol. You are running a 4-to-6-week mini-stability check to catch obvious failure modes before you approve production:
Cycle sample at 40°C constant for 4 weeks. Pull aliquots at week 1, week 2, week 4. Check organoleptic, pH, viscosity at each pull.
Run a freeze-thaw cycle: minus-5°C for 24 hours, room temperature for 24 hours, three cycles. Emulsion separation, crystallization at the bottle wall, or visible cloudiness are failure indicators.
Run a light-stress test: place a sample in a south-facing window for 2 weeks. Color shift, oxidation odor, or active degradation are failure indicators. This catches photo-unstable actives like vitamin C derivatives, retinol, and certain plant phenolics.
Ask the Korean ODM for their internal 4-to-6-week stability table at the time the sample ships. If they have already run this and the data is clean, the founder-side mini-stability becomes a confirmation rather than a discovery exercise.
Step 5: Preservative Efficacy & Microbiology (When to Run, When to Wait)
Preservative efficacy testing under ISO 11930:2019 is the cosmetic-industry challenge test. The lab inoculates the product with five microorganisms (Pseudomonas aeruginosa, Staphylococcus aureus, Escherichia coli, Candida albicans, Aspergillus brasiliensis), then counts microbial recovery at 7, 14, and 28 days. The test typically runs 4-6 weeks from sample arrival to report, and costs roughly USD 450-800 per sample at US contract labs.
For first-round sample evaluation, founders generally should not commission their own preservative efficacy test. Here is the logic:
A Korean ODM with ISO 22716 cosmetic GMP certification runs an internal preservative efficacy panel against their preservative system at formulation time. Ask for that data sheet. If they have it, the first-round sample is already screened.
An independent preservative efficacy test makes sense at the pre-production milestone, when the formula has been frozen and you are validating the production-equivalent batch. That is a 4-6 week timeline that you want to lock in before pilot batch is scheduled, not now. (For where sample evaluation sits in the 6-month launch timeline, see our launch timeline playbook.)
The first-round sample is too early. Reformulation is still on the table. Paying for a preservative efficacy test on a sample you might never produce is wasted budget.
The exception is products with high-risk substrates: water-rich formulations without traditional preservatives, products targeting eye area or for pediatric use (where ISO 17516:2014 sets the microbiological limit at 100 CFU/g instead of the standard 1,000 CFU/g for leave-on products), and any formula with botanical or fermented extracts in concentrations above 1 percent. For those, run a preservative efficacy test at first round, because reformulation cycles for preservation are long.
Step 6: Cross-Test Against Your Brief
Pull out the original brief you sent the Korean ODM. Read it line by line. For each specification in the brief, write "match" or "miss" next to the sample column on your evaluation sheet. Specifications that commonly drift in a first round:
Target finish (dewy, semi-matte, matte) versus delivered finish
Spreadability and slip versus brief
Hero active concentration on the INCI list versus the brief number
Fragrance load and family
Hero claim ingredient placement on the INCI list (top-third versus bottom-third)
Color and clarity
If the brief asked for a 5 percent niacinamide essence and the INCI list shows niacinamide in position seven, the actual concentration is more likely 2-3 percent. Ask the ODM to confirm the in-batch concentration with an HPLC or UV-Vis assay number. Korean ODMs run these routinely on hero actives. The number should match your brief within plus-or-minus 10 percent or you have a different product than you ordered.
This is the step where most "the sample is close, but not quite right" feelings get resolved into specific, actionable revision requests. Send the Korean ODM a single document with the specification, the delivered value, and the requested change, with a target revision date.
I'm Liz, I run ALTA MEET from Manhattan, NYC. Sample evaluation is the place where the most expensive mistakes hide, because the cost of an incorrect approval shows up 12 weeks later as a 1,000-unit production batch that does not match what you promised your customer. If you want a quick gut-check on a sample round you're sitting on, grab 15 minutes free with me and walk through it before you write back to your ODM.
Step 7: Wearer Trial Protocol (Small, Structured, Honest)
A first-round wearer trial is not a clinical study. It is a structured 5-to-7-day use-test by 3-5 people who reflect your target customer. Run it after Steps 1-6 only on candidates that have passed the basic checks. The point is to detect failures that the technical data cannot show: a fragrance that becomes cloying after 30 minutes of wear, a finish that pills under a sunscreen, a feel that is fine in the hand but unpleasant on the face.
Protocol skeleton:
3-5 evaluators, each receives the candidate in an unlabeled vial coded A, B, or C
5-7 day use, twice daily, applied to half the face if you have multiple candidates to compare
Daily 1-paragraph note: pickup, application, immediate finish, finish at 2 hours, finish at end of day, any irritation or transient redness
A final 1-page summary with a clear vote: this is the one, send back for revision, or do not produce
Wearer compensation: USD 20-50 per person in product or gift card is normal and ethical
Document irritation rigorously. If even one evaluator reports stinging on the eye area for a product not designed for eye area, that is a meaningful signal. If multiple report it, the product needs to be reformulated or repositioned before you go to production.
Step 8: Cost-of-Revision Math (The Decision Anchor)
Once Steps 1-7 are complete you have a defensible technical and experiential view of the sample. The decision is rarely "perfect." It is usually "good enough to produce" versus "another round." The cost-of-revision math sets the bar.
A second-round revision from a Korean ODM typically:
Costs USD 300-800 in additional sample-batch and shipping fees (some Korean ODMs comp the second round, especially on larger MOQs; ask explicitly)
Adds 3-4 weeks of timeline (formulation chemist queue, mini-stability re-confirm, sample shipping)
Compounds across rounds. Round 3 and Round 4 are common when the brief was vague or the first revision was not specific.
If the candidate is at 85-90 percent of brief and the gap is in a non-critical attribute (a slightly stronger fragrance, a marginally faster sink-in time), the math often favors approving with a documented in-batch tweak at the production lot rather than running a full second round. If the gap is in a hero claim, a structural ingredient, or anything that touches the marketing story or regulatory label, run another round.
What ALTA MEET has seen across roughly two dozen indie K-beauty briefs: founders who run two well-structured rounds beat founders who run four loose rounds, every time. The variable is brief specificity and sample evaluation discipline, not the number of samples.
Step 9: Sign-Off Decision Tree
When all the data is in, the approval decision should fit one of three buckets. Pick one in writing and communicate it to the Korean ODM the same day.
Approve for production: candidate matches the brief in all critical attributes (hero active concentration, pH, viscosity, base finish, safety profile). Non-critical drift is documented in the in-batch tweak request. Production go-ahead is contingent on a confirmed pre-production sample at the pilot batch scale that re-validates the approved profile within plus-or-minus 5 percent on each measured attribute.
Request a second round: critical attribute drift, hero claim concentration miss, or any safety or stability flag. Send the Korean ODM a single document with the gap, the requested change, and a target revision date. Do not ask for "lighter" or "better." Ask for measurable changes against the original brief.
Reject and re-brief: candidate is structurally different from what you described. This is rare and usually points to a misunderstood brief. Re-issue the brief with more specificity (target reference product, target finish profile, target measurable values). A second ODM bid at this stage is often worth the time.
Whichever bucket you pick, save the evaluation log, the photos, the wearer trial summary, and the cost-of-revision math in a single folder named with the candidate code and date. This folder is the contemporaneous record that, two years from now, lets you understand why a successful product is successful and why a failed reformulation went sideways.
Common Mistakes Indie Founders Make on Sample Evaluation
Skipping the tech sheet. Founders smell, feel, and decide. The tech sheet has the numbers that make the decision real.
Conflating "I love this" with "this matches the brief." A founder's personal preference is one signal; the brief is the contractual reference.
Running a single-day evaluation. Wear is temporal. A sample that feels great in minute one and unpleasant at hour four is a different product than a sample that feels good at both points.
Asking for "softer," "fresher," or "more luxe." Korean ODMs work in numbers and reference products. Translate vague feedback into measurable changes before sending it.
Approving without a pre-production sample. The first sample was a lab batch. The first production batch is a 1,000-unit run with scale-up variables that can change viscosity, pH, and finish. A pre-production sample at pilot scale is the bridge.
Treating preservative and stability data as optional. The Korean ODM has it; ask for it. If they cannot produce it, that is a signal about the ODM itself.
A Realistic Phased Timeline for Sample Rounds
Sample shipment from Korea to US 4-7 — DHL or FedEx international air, customs clearance
Steps 1-3: unboxing, organoleptic, pH and viscosity 1-2 — Day 1 inspection, Day 2 confirmation read
Step 4: mini-stability (4 weeks) 28-30 — Parallel to other steps; reads at weeks 1, 2, 4
Steps 5-6: tech sheet cross-check and brief gap analysis 1-2 — Concentrated on day 2-3 after arrival
Step 7: wearer trial 5-7 — Started after Steps 1-6 pass; before Step 8 decision
Step 8-9: cost-of-revision math and decision 1 — Single working session
Total first-round evaluation 5-6 weeks — Drives directly into pre-production sample request
If a Korean ODM asks for an approval decision in under two weeks, the request is unrealistic for any candidate that has not already been through a structured prior round with the same brand. Push back. The cost of a delayed approval is small; the cost of an undercooked approval is the production batch.
FAQ
Q1. How many candidates should a Korean ODM ship in a first round? Two to three is standard for a single SKU. Five or more suggests the brief was too open and the ODM is hedging. If you receive five candidates, reset the brief before evaluating: pick the two closest to your intent and run the protocol on those.
Q2. Should I send the samples to a third-party lab in the US for confirmation testing? At first round, usually no. Third-party pH and viscosity confirmation runs USD 80-150 per sample, and is a reasonable spot-check when the ODM tech sheet looks unusual. Full third-party stability and preservative efficacy testing belong at the pre-production milestone, not at first round.
Q3. What if the sample arrives separated or cloudy? Stop the round. Photograph the vial as received, contact the Korean ODM with the photo, ask whether the formulation is meant to be shaken before use (some emulsions are; most are not). If it is not, request a re-ship. Do not run the rest of the protocol on a compromised sample.
Q4. How do I verify a hero active concentration without paying for HPLC? Ask the Korean ODM for the in-batch HPLC or UV-Vis assay number. Korean ODM labs run these as part of standard QC. If they refuse or cannot produce the number, that is a meaningful signal about lab capability. Third-party HPLC for niacinamide, vitamin C, or peptides typically costs USD 250-450 per assay in the US.
Q5. What does ISO 22716 cosmetic GMP have to do with sample evaluation? ISO 22716 is the cosmetic GMP standard most reputable Korean ODMs operate under. It governs how the ODM documents formulation, batches, and QC, including the data they should be able to produce on every sample they ship. If the ODM is ISO 22716 certified, the tech sheet, stability table, and preservative panel should all be available by default. If they are not, ask why.
Q6. Should I run a wearer trial before or after the technical evaluation? After. A wearer trial on a sample with the wrong pH, separated emulsion, or off-spec active concentration is wasted. Technical screening (Steps 1-6) eliminates fundamentally flawed candidates so the wearer trial is reading real product, not a defective lab batch.
Q7. The Korean ODM is pushing a fast approval. Is that normal? Sometimes a Korean ODM has tight raw material lead times and is anchoring on a real production schedule. Sometimes it is internal Korean ODM pipeline pressure that has nothing to do with your timeline. Ask which it is. If it is the first, work to the constraint with eyes open; if it is the second, hold the line on the evaluation protocol.
Key Takeaways
Sample evaluation is the gap step between briefing and production where most indie K-beauty founders lose 6-12 weeks of timeline and significant reformulation budget.
The 9-step protocol covers unboxing, organoleptic, pH and viscosity, mini-stability, microbiology decisions, brief cross-check, wearer trial, cost-of-revision math, and sign-off.
Korean ODMs operating under ISO 22716 should produce tech sheets, internal stability data, and preservative efficacy panels by default; if they cannot, that is a signal about the ODM, not about your sample.
The cost-of-revision math is the decision anchor. Two well-structured rounds beat four loose ones.
A pre-production sample at pilot scale is non-negotiable before approving a 1,000-unit production batch.
Working With ALTA MEET
ALTA MEET is a Manhattan, NYC-based K-beauty consultancy. We sit in your sample-evaluation seat: we run the protocol above on your behalf, translate your feedback into specifications the Korean ODM will act on, and make sure the pre-production sample matches the round you approved. We work with first-time indie founders and growing US K-beauty brands across hubs in NYC and Seoul.
If you have a sample round on your desk and you want a second pair of eyes before you reply to the Korean ODM, book your free 15-min K-Beauty manufacturing gut-check with Liz. Bring the tech sheet, the brief, and a photo of the vials. We will walk through it together.
Reviewed for accuracy by ALTA MEET's formulation consulting team.
