Oily skin: 38.9% by questionnaire, 11.5% by meter
Most brands size their oily-skin line from consumer research. Measure the same people with an instrument and the oily share drops below a third of that. Skin-type labels follow the yardstick, not the skin, and running a second survey does not close the gap.
Skin type depends on which yardstick you use. In a Korean study that typed the same 131 women four different ways, a questionnaire put 38.9% in the oily bucket while facial sebum measurement put 11.5% there. Self-assessment and the questionnaire agreed closely with each other (r=0.709) but tracked the instrument only weakly (r=0.287 and r=0.393).
Which yardstick produced your oily-skin number?
The decision to widen an oily-skin line usually comes out of consumer research. Measure the same people with an instrument and the oily share falls below a third of the survey figure. Before you size the line, check which yardstick produced the number in your hands.
One Korean study typed the same people four ways at once. Rou and Park measured 131 female university students (mean age 19.9) over May and June 2003, tabulating facial sebum, facial hydration, a questionnaire and the subject's own self-assessment separately. It is a single cohort from twenty years ago, so the percentages are not national figures. What the design does show, and almost no other dataset does, is how much the yardstick moves the answer.
| Yardstick | Dry | Normal | Oily | Combination |
|---|---|---|---|---|
| Facial sebum (instrument) | 33.6% | 55.0% | 11.5% | no bucket |
| Facial hydration (instrument) | 47.3% | 52.7% | 0% | no bucket |
| Questionnaire | 16.1% | 22.1% | 38.9% | 22.9% |
| Self-assessment | 22.1% | 13.7% | 17.6% | 46.6% |
The questionnaire put 38.9% in the oily bucket; the sebum meter put 11.5% there. Combination runs the other way: it is the largest bucket under self-assessment at 46.6%, and instrument typing has no combination bucket at all. That is why a brief asking for a "combination skin" product is hard for a factory to turn into a formula.
The surveys agree with each other, not with the meter
The direction of the disagreement matters more than its size. Self-assessment and the questionnaire correlated at r=0.709 in the same study. Ask a consumer the same thing two ways and you get much the same answer back.
Against the instrument, those two sat at r=0.287 and r=0.393. A second survey does not pull the estimate toward the measurement. Two consumer datasets agreeing with each other is not evidence that either one matches what an instrument reads.
In practice the gap turns into inventory. Size an oily SKU mix off 38.9% and you order roughly three times the demand the measurement implies. Dry runs the opposite way: hydration typing put 47.3% in the dry bucket against 16.1% on the questionnaire, so the moisturising line gets built too small.
The instrument bands are a human choice too
None of this makes instrument typing the true answer. Someone still picks the cut-offs, and the study used these.
| Axis | Dry | Normal | Oily |
|---|---|---|---|
| Sebum | under 10 AU | 10-30 AU | 30 AU and above |
| Hydration | under 40% | 40-60% | 60% and above |
The two axes do not explain one another. Facial sebum and facial hydration correlated at r=-0.090, which is not significant, and 12 of the 15 people typed oily on sebum fell into the dry bucket on hydration. The consumer sentence "oily but tight" shows up in the measurements as well.
Conditions travel with the numbers. The readings were taken indoors at 24±1°C and 60±2% relative humidity, and the typing used values from 12 hours after cleansing. Change those and the same face lands in a different bucket, so when you request measurement data ask for the timing and room conditions alongside the instrument name.
The label risk sits next to the type, not in it
"For dry skin" normally reads as a usage note. The exposure is the efficacy sentence beside it. Under the Cosmetics Act, where the MFDS commissioner deems a claim to need substantiation, the company has 15 days from the request to file, and continuing to advertise without filing brings a stop order - with up to one year in prison or a fine of up to 10 million won for ignoring it.
What counts as substantiation is set out in three lines in Article 23(2) of the Enforcement Rule.
| Type | Requirement |
|---|---|
| Test results | Human application test data, in-vitro test data, or survey data of equivalent or higher standing |
| Survey results | Sampling, questions and method consistent with the survey's purpose and its statistics |
| Method | Scientific and objective, academically established or generally accepted in the industry |
Consumer research lives on the second line. Submitting only a general consumer survey behind an efficacy claim, though, is listed among the cases that are not accepted. For human application testing the operator is part of the requirement: the Regulation on Substantiation of Cosmetic Labelling and Advertising (MFDS Notice 2020-80) requires the work to run under a specialist physician or hospital, or someone with five or more years of cosmetic human-application-testing experience at a university or recognised research institute.
Some wording is closed off outright. An MFDS written reply from December 2024 notes that Annex 1 of the labelling and advertising guideline lists "relieves itching" as a prohibited expression, while carving out wording that says moisturisation helps temporarily relieve itching caused by dry skin. One clause separates the two.
Three lines to settle before the order
First, state whether the type mix comes from survey or from measurement, and use one of them. Mixing the two is what makes production volumes wobble. Second, keep the type label and the efficacy sentence on separate lines of the brief, because the substantiation burden attaches to the second. Third, if measurement data is coming, specify instrument, timing and room conditions in the request.
Products made in Korea for export only, with no domestic sale, can follow the destination country's rules under Article 30 of the Cosmetics Act. Translating the Korean product page as it stands still runs into the destination's own standards. You can see which ingredients are moving on the shelf right now in today's ingredient signal, and price a type-by-type SKU mix in the product planner.
«Summary»
- In the same 131 people, the oily share was 38.9% by questionnaire and 11.5% by sebum meter - more than a threefold gap.
- Self-assessment and questionnaire agree at r=0.709; both sit near r=0.3 against the instrument. More surveys will not narrow it.
- Sebum and hydration are independent axes (r=-0.090). Twelve of the 15 oily-by-sebum subjects were dry by hydration.
- Substantiation risk comes from the efficacy sentence beside the type label, not from the label. Once requested, the clock is 15 days.
*Sources: Rou F. and Park Y., "Comparison of determined skin types by different factors of facial skin hydration, sebum content and surface pH levels", Journal of the Korean Society of Cosmetology Education, pp.67-81 (survey May-June 2003, 131 female students) - PDF / substantiation scope and filing deadline - Labelling and advertising substantiation, Korea Easy Law / notice number and effective date - Regulation on Substantiation of Cosmetic Labelling and Advertising, Korean Law Information Center / test operator requirement and rejected cases - Cosmorning, substantiation regulation amendment / prohibited expression and its carve-out - MFDS written reply, Korean Law Information Center, December 2024 / skin-barrier study design - THE K BEAUTY SCIENCE, 17 June 2021. Accessed 12 November 2026.*
Korean private label skincare for this skin concern
A concern like skin type is where Korean private label skincare is strongest: the base formulas already exist and have been tested on Korean shelves. We adapt texture, actives and claim wording to your market, and start from a low MOQ so the first run is a test, not a bet.
Korean cosmetics private label — cost, MOQ and timeline →
Start a sample request · Browse ready formulations
Questions people ask
The two numbers differ by more than three times in the same group - 38.9% oily by questionnaire against 11.5% by sebum meter. Use the survey for purchase intent, and set SKU mix and formulation from measured data.
It does not move toward the instrument. Self-assessment and the questionnaire correlated at r=0.709, yet both sat near r=0.3 against sebum measurement. Two consumer surveys agreeing with each other is not evidence that either matches a measurement.
No. The study cut sebum and hydration into separate bands, which is why 12 of the 15 people typed oily on sebum landed in the dry bucket on hydration.
The efficacy sentence next to it is the exposed part. Where the MFDS commissioner deems substantiation necessary, the company has 15 days from the request to file, and failing to file brings an order to stop the advertising.
The MFDS skin-barrier guideline issued in June 2021 asks for at least 30 valid results in each of the test and control arms, and a design of at least four weeks to allow for stratum corneum recovery.