Screening by review count hides the category leader
If you have ever shortlisted Korean products by review count, the ordering runs the wrong way. In our daily Olive Young category top 30, ranks 1 to 10 carry a median of 42 reviews while ranks 21 to 30 carry 288. Thirty-five of 41 categories point the same way.
Review count does not explain rank. In the Olive Young category top 30 collected 5 November 2026, ranks 1 to 10 carry a median of 42 reviews against 288 for ranks 21 to 30, so the heavier review counts sit lower. Across the 41 category leaders, the median review count is zero.
Ranks 1 to 10 carry a median of 42 reviews
Review count is the filter overseas buyers reach for first when shortlisting Korean products. A heavy review count reads as validation already earned. In our daily Olive Young category top 30, collected 5 November 2026, that ordering runs backwards.
Splitting the 1,209 slots across 41 categories into rank bands, the top ten carry a median of 42 reviews. Ranks 11 to 20 carry 150, and ranks 21 to 30 carry 288 — the count rises as the rank falls. Of those top ten slots, 57.1% sit below 100 reviews, and 89 of the 410 have no reviews at all.
The mean tells a different story. Ranks 1 to 10 average 2,788 reviews against 2,349 for ranks 21 to 30. A handful of long-running products mixed into the top drags that average up, while the middle of the distribution has moved to newly launched items.
| Rank band | Slots | Median reviews | Under 100 reviews | Over 1,000 reviews |
|---|---|---|---|---|
| 1–10 | 410 | 42 | 57.1% | 20.7% |
| 11–20 | 405 | 150 | 44.7% | 26.2% |
| 21–30 | 394 | 288 | 36.3% | 30.5% |
The column to read is the third one. Median review counts differ almost sevenfold between the top ten slots and the bottom ten, and the heavier side is the bottom.
Thirty-five of 41 categories point the same way
A pattern in one or two categories would not be worth acting on. So we calculated the rank correlation (Spearman) between position and review count separately for each category.
Thirty-five of the 41 return a positive coefficient. A higher rank number — that is, a lower position — comes with more reviews. Pooled across all slots the figure is +0.204, and cream and foot care come in strongest at +0.61.
Only six categories run the other way. Styling sits at -0.31 and eye makeup at -0.27, with heavily reviewed items holding the upper slots. Colour and hair are the aisles where repeat purchase is tied to personal preference, and only there does review count explain any of the ranking.
View as table
| Cream | +0.61 |
|---|---|
| Foot care | +0.61 |
| Hair treatment | +0.56 |
| Hair removal | +0.55 |
| Tanning & after-sun | +0.52 |
| Peeling & scrub | -0.15 |
| Eye makeup | -0.27 |
| Styling | -0.31 |
Thirty of 41 leaders sit below 50 reviews
Pulling out the number one slot alone makes the point sharper than the bands do. Across the 41 category leaders the median review count is zero, and 30 of them sit below 50 reviews.
Counting the items flagged as new entries explains why. On the same day, 115 items entered the top 30 across 30 categories, carrying a median of two reviews between them. Fully 73.9% sit below 100 reviews and are holding a top 30 slot regardless.
Olive Young category rank tracks recent selling speed, while reviews attach well after the sales have accumulated. Filter on review count and what survives the filter is not what sells now, but what sold heavily last year.
| Category | Rank | Brand | Reviews | Shelf price |
|---|---|---|---|---|
| Essence/serum/ampoule | 1 | MilyMily | 0 | ₩20,900 |
| Cream | 1 | HERA | 0 | ₩45,600 |
| Sunscreen | 1 | THE WHOO | 6 | ₩15,600 |
| Cream | 26 | AESTURA | 45,064 | ₩43,600 |
| Nose pack | 28 | HATHERINE | 25,830 | ₩6,000 |
The top three rows and the bottom two come from the same shelf on the same day. A cream with 45,064 reviews sits at rank 26, and a cream with none sits at rank 1. Today's figures from the same tracking sit on today's ingredient signal.
Star ratings separate even less
Reaching for the star rating instead looks like the fix, and it is flatter still. Ranks 1 to 10 show a median rating of 4.80, and ranks 21 to 30 show 4.80 as well.
Korean retail ratings bunch almost entirely into the 4.7 to 4.9 band, which leaves nothing to separate thirty slots with. Review count at least retains a direction, even an inverted one; the rating has no direction at all.
What to read instead
None of this argues for discarding reviews. Review count tells you how long something has been selling, and it misleads the moment it is asked whether something sells now. The two questions need separate columns.
Whether it sells now comes from rank itself, the new-entry flag, and how many days the position holds. How long it has sold comes from review count and the rate at which reviews accrue. Put both axes together and "few reviews, high rank" separates cleanly from "many reviews, low rank".
In practice that split changes order size. The first group is safer taken in small volume and reviewed after two or three months of rank data; the second offers more room on price but less freshness. To lay rank, reviews and price side by side per category, work it through the product planner.
In short
- Ranks 1 to 10 carry a median of 42 reviews against 288 for ranks 21 to 30, as of 5 November 2026.
- Thirty-five of 41 categories show a positive rank-to-review correlation, and the pooled figure is +0.204.
- The 41 category leaders carry a median of zero reviews, and 30 of them sit below 50.
- The 115 items flagged as new entries carry a median of two reviews, with 73.9% below 100.
- Median star ratings are 4.80 at both the top and the bottom of the thirty, so they separate nothing.
*Source: our daily Olive Young category top 30 tracking, collected 5 November 2026 at 09:40 KST. Rank-band figures cover the 1,209 slots carrying a recorded review count; correlations are Spearman, calculated per category across the 41 categories with at least 15 recorded slots. New-entry flags and star ratings are as recorded in the same tracking. Tracking window 16 July to 4 November 2026 (102 days).*
Turning this signal into a Korean private label product
A signal is only useful if you can ship before it fades. Korean private label manufacturing lets you go from a ready formula to a first batch in weeks, at a low MOQ, with the cost known up front.
Korean cosmetics private label — cost, MOQ and timeline →
Start a sample request · Browse ready formulations
Questions people ask
The opposite. In the same tracking, ranks 1 to 10 carry a median of 42 reviews while ranks 21 to 30 carry 288 — close to a sevenfold gap. Of the top ten slots, 57.1% sit below 100 reviews.
Calculated per category, 35 of 41 return a positive rank correlation between position and review count. Pooled, the figure is +0.204, with cream and foot care strongest at +0.61.
The 41 leaders carry a median of zero reviews, and 30 of them sit below 50. Rank tracks recent selling speed, while reviews accumulate afterwards.
It answers how long something has been selling. Whether it is selling now comes from rank itself and the new-entry flag, and separating the two axes splits newly launched test items from late-life ones.