Olive Young, Korea's biggest beauty retailer, runs two stores. The Korean one (oliveyoung.co.kr) sells to people in Korea. The Global one (global.oliveyoung.com) ships abroad. A product that is on both has two separate review sections, written by two different sets of buyers.
I wanted to see how different those two crowds are, so on October 10 I took one product and pulled the newest 500 reviews from each store.
The product
SKIN1004 Madagascar Centella Hyalu-Cica Water-Fit Sun Serum, one of the best-known Korean sunscreens abroad.
- Korean store: 17,292 reviews, average 4.9
- Global store: 11,963 reviews, average 4.9
I checked that it is the same item and not just a similar name. The Global listing is the twin pack with barcode 8809913830122. On the Korean store, every review carries the barcode of the exact option the reviewer bought, and that same barcode is there, next to the single tube and a gift set that share the listing.
Same sunscreen, same 4.9. Everything else is different.
Korea reviews it five times faster
| Korean store | Global store | |
|---|---|---|
| Newest 500 reviews go back to | Sep 29 | Aug 7 |
| Days covered | 12 | 65 |
| Reviews per day | 41.7 | 7.7 |
The Korean store collects 5.4 times as many reviews per day from one country as the Global store does from all the others. And it is October, not sunscreen season.
Part of the reason is in the data: 163 of the 500 Korean reviews (32.6%) are marked as in-store purchases. People buy it at one of Olive Young's physical shops and review it in the app later. The Global store has no shops to feed it.
The stars are the same
| Korean store | Global store | |
|---|---|---|
| 5 stars | 455 (91.0%) | 453 (90.6%) |
| 4 stars | 31 | 41 |
| 3 stars or lower | 14 | 6 |
If you only look at the rating, the two markets agree almost to the review. So the rating is the least useful column here.
The world posts photos, Korea doesn't
- Global store: 38.6% of reviews have a photo
- Korean store: 16.4%
International buyers also fill in their skin profile far more often: 342 of 500 (68.4%) against 142 of 500 (28.4%) in Korea.
Two stores, two ideas of "skin tone"
Both stores ask reviewers for a skin tone. They do not mean the same thing by it.
The Korean store asks for your personal colour, the seasonal system that is everywhere in Korean beauty: cool, warm, spring warm, summer cool, autumn warm, winter cool. It is about undertone, not depth.
The Global store asks how light or dark you are: porcelain, fair/light, medium, tan, olive, deep, dark, ebony. Of the 339 reviewers who answered:
| Skin tone | Reviewers |
|---|---|
| Fair/Light | 155 |
| Medium | 98 |
| Porcelain | 45 |
| Tan | 21 |
| Olive | 9 |
| Deep | 8 |
| Dark | 2 |
| Ebony | 1 |
For a sunscreen this matters. "No white cast" from a porcelain reviewer and from a deep-skinned reviewer are not the same claim, and only the Global store lets you tell them apart. In this sample you would have 41 reviews from tan-to-ebony skin to check it against.
Skin type is nearly identical. Concerns are not.
Skin type, among reviewers who gave one:
| Korean store (142) | Global store (342) | |
|---|---|---|
| Combination | 40.8% | 42.7% |
| Dry | 25.4% | 20.2% |
| Oily | 16.9% | 11.1% |
| Sensitive | 12.0% | 15.2% |
Close enough that the product clearly sells to the same kind of skin everywhere. What those people worry about is another story. Top five skin concerns:
| Korean store (129 reviewers) | Global store (335 reviewers) | ||
|---|---|---|---|
| Dark spots | 36.4% | Blackheads | 47.5% |
| Brightening | 27.9% | Acne | 41.8% |
| Breakouts | 27.1% | Brightening | 40.9% |
| Flakiness | 24.8% | Moisturising | 38.5% |
| Pores | 19.4% | Dullness | 27.2% |
The Korean buyer of this sunscreen is thinking about pigmentation. The international buyer is thinking about blackheads and acne. Same tube, and a marketer would write two different product pages for it.
(The two stores use different lists of concerns, so compare the rankings, not the exact percentages.)
A quarter of the "global" reviews are in Japanese
I expected the Global store to be English. I sorted the 500 reviews by the script they are written in:
- Latin script (English, Spanish and so on): 341
- Japanese (contains kana): 117, or 23.4%
- Chinese characters only: 20
- Cyrillic: 13
- Korean: 7
- Arabic: 2
This is a rough split by character set, not real language detection. But the Japanese share is hard to miss, and it is the kind of thing a "global" average hides.
What only one store can tell you
Global store: a separate score for sun protection. Each reviewer rates attributes as well as the product. Texture averages 4.88 and Sun protection 4.79. Of the 453 people who gave the product five stars, 51 gave its sun protection less than five. They love how it feels and are not fully sure it protects. You cannot see that in the star rating.
Korean store: how long people used it. 285 of the 500 reviews (57.0%) carry Olive Young's flag for a review written after a month or more of use. The Global store has nothing like it.
Limits
- One product, the newest 500 reviews per store, collected on one day.
- The Korean store stops serving reviews after the first 500 of each sort order unless you log in. I did not log in, so 500 is the ceiling per sort there. Sorting another way gives a different 500. The Global store kept going to at least 1,300 when I tested it.
- Skin profiles are optional and self-reported.
How I got the data
Both review sections load from JSON endpoints, so there is no HTML to parse. I used my own Olive Young Reviews Scraper on Apify, which now takes product numbers from either store in the same run:
{
"products": ["A000000215559", "GA230518746"],
"sortBy": "DATETIME_DESC",
"maxReviewsPerProduct": 500
}
A… numbers are the Korean store, GA… numbers are the Global store. Each row comes back with the same core fields (rating, text, date, photos, reviewer skin profile) plus whatever is specific to that store.
If you have a product you would like compared across the two stores, tell me which one in the comments.
Top comments (0)