Industry

Across 2,372 Korean singers, rappers, composers and music groups, the top 24 (1%) account for 21.5% of a month's English Wikipedia reading. The bottom half — 1,186 names — account for 3.7%.

Among 774 Korean music acts with enough monthly English Wikipedia history to measure, 731 have already fallen to half their peak reading — and the typical group got there in 2 months.

NewJeans, IVE, LE SSERAFIM, aespa, ITZY, (G)I-DLE, STAYC and NMIXX all averaged fewer English Wikipedia reads a month in the last 12 months than in their second year after debut. The declines range from 9% to 53%.

Among 9,249 Korean public figures, February is the most common birth month, June the least — a 10,000-draw test found this is essentially not chance. Narrowed to 1,317 idols, the same test can't tell the pattern from noise.

Across 269 weeks of Netflix's Top 10 charts in 93 tracked markets, 974 Korean titles have placed at least once. 447 of them — 45.9% — never placed top 10 anywhere but a single country. The median title charted in just 2.

Only 12.6% of 3,671 Korean music credits and 14% of 4,554 acting credits have measurable English Wikipedia reading. Of 992 people credited with both, 47.5% do — and exist in a median of 10 language editions against 1.

In 14 of 51 comic-to-screen adaptations we measured, the original comic reads more on English Wikipedia than the film or show made from it. Three of those 14 are different Garfield adaptations — all losing to the same original.

The Esports at the 2026 Asian Games page averaged 276 English Wikipedia reads a day in early September. By the day after the opening ceremony, that hit 2,537 — 9.2 times the baseline, building for days before the Games opened.

Among groups with a recorded end date on Wikidata, Korean groups' median lifespan is 3 years against 5-7 for Japan, the US and the UK. Not a ranking — the Korean cohort was founded a median 31 years later than the UK's.

Korean cast in 7+ Netflix Top 10 titles had a narrower 30-day Wikipedia reading spread (0.25) than those in just 1 title (0.31). A resample test says that gap shows by chance alone 11.2% of the time — short of what we call confirmed.

We scanned the 30 August 2026 daily top-1,000 most-read English Wikipedia articles for Korean film and TV titles. Two made the list: Mousetrap at rank 186 (21,033 reads) and Four Hands, Two Sonatas at rank 340 (15,420 reads).

A Seoul court suspended EXO-CBX's contract with its agency on 15 September. We can't check that day's reading yet, but an earlier filing story moved all three pages together — a bigger, unexplained rise the day before hit only Baekhyun.

We photograph the front pages of Korean forums and English-language Korea subreddits once a day. Over eight consecutive days, 0.2 percent of Korean forum posts were still there the next evening. In the subreddits it was 36 percent.

We counted quoted ages in Korean front pages last night, then won amounts in the same 1,365 headlines. The business desk uses money in 22 percent and ages in none. The entertainment desk is the mirror image. No headline carried both.

We keep a daily copy of three English-language Google News queries about Korea. Over nine days the phrase 'KPop Demon Hunters' appears in 645 of 1,110 K-pop items. The busiest subject in the drama feed reaches 71, in the economy feed 90.

Ten days of Korean front pages, 1,365 headlines. One entertainment desk opened 27 percent of its stories with an age in quote marks. The other did it in 3 percent. The general and business papers: zero in 1,080.

A forum called spreading visitors beyond the capital the unfinished task. In the admissions table 58 percent of foreign entries are already outside it, and the top district is Gangneung, not Jongno. But 112 of 206 districts count nobody.

If you cannot read Korean, the only legible part of a front page is its Latin script. We counted every Latin word in 1,388 headline-days. The K-culture desk prints them four times as often as the general paper, and prints a different kind.

Jin has topped a Korean idol poll for 68 straight weeks. On English Wikipedia lookups he is fifth of seven BTS members. But ranks three to six sit inside 7% of each other, so the ranking barely means anything in the middle.

Every person named in one day of two Korean entertainment desks, against how many the English-speaking world looks up. Sixteen of eighteen are findable. The day skews to actors two to one, but per head they lead idols by just 1.13 times.

Virtual performers have been shipping for 27 years. Over twelve months the biggest virtual page beat six real K-pop groups combined — by 0.4%. Move the window one month and the lead disappears. Last month the real groups won by 3.2 times.

HYBE sold 17.5% more than the year before and still reported a net loss 85 times larger. SM, JYP and YG all grew revenue and profit. Filed figures, in English, with the filing receipts named.

Jennie's solo single matched a BTS record, so we counted lookups for three groups and all fifteen members. Only Lisa passed her own group page — by 7,413 out of 1.8 million. Added up, BLACKPINK's four beat its name 3.25 times.

Cube, SM and Hybe all had something out this week. For each of nine K-pop companies we hold a reader count, but the share of the roster we can see runs 37% to 67% — so the ranking everyone wants is the one thing the data will not give.

Wikipedia publishes each country's most-read articles daily. Across 20 countries and 14 days, 87 Korean articles ever appeared — 63 of the 93 appearances were Japan alone. Nine countries had none.

Korea runs a public platform holding 1,344 culture datasets — tourism, leisure, K-pop audiences. We counted every listing. Most cost nothing. Only 322 have been updated in the last twelve months.

Seoul Institute of the Arts has 352 people in our roster, more than any school. Twenty-eight are in a group. School of Performing Arts Seoul has 102.

Seoul Institute of the Arts has more alumni in our Netflix set than any school. Their titles reach a median of 16 countries. Dongguk's reach 41.5.

Kim Sang-sik's Vietnam beat Thailand 4-2 on aggregate in Hanoi. We do not count football. What we count is 268 weeks of Netflix country lists — and by that ruler the two finalists are not the same country at all.

Wikidata records 691 Korean musical groups with a formation year. The 2023-25 average is 31% of the 2015-17 average. Then we ran the same count for Japan, the United States and the United Kingdom — and all three fell too.

Reads of Korean travel articles in four Southeast Asian Wikipedias peak in December; flights on the Asia routes peak in January. The steadier signal is the other end — both fall to their lowest point in the same month.

Twenty-six Korean titles could be timed from their peak across four Southeast Asian Wikipedias. The median fell below half in two months, nine of them in a single month, and nineteen later rose above half again.

Removing a single title left one of our medians exactly where it was and moved the other by 89% of itself. The interquartile range rated the two almost identically. We ran neither before publishing.

Of 59 Korean titles fetched across four Southeast Asian Wikipedias, nine could be measured before and after their peak. Twenty-two were disqualified because their article did not exist before the title arrived.

Reads of Korean travel articles in four Southeast Asian Wikipedias fell 30% in a year. Korean culture articles fell 14%, Japanese and Taiwanese culture articles rose, and air passengers on the Asia routes barely moved.

We tried to set what Southeast Asia reads about Seoul beside where foreign visitors walk. In 16 of 24 districts the admissions table counts no tourist site at all — Jung, which holds Myeongdong, among them. It maps where counting stops.

We measured 2,238 places in Korea against four Southeast Asian Wikipedias. YG Entertainment scores 111.47 and Seoul 106.54. Twelve companies out-read 520 cities and districts on the median. What this cannot see is a single restaurant.

Split 22 brands by kind and the four Southeast Asian countries swap places. Indonesia is first among car makers on 53.79 reads per million and last among luxury houses on 38.45. A single ranking hides that.

Groups, solo musicians, actors, athletes — the Malay Wikipedia takes 7.6% to 8.6% of the four-country reading for each. For luxury and car brands it takes 23%. Four categories in a narrow band, one well outside it.

Two works can carry the same English title, and a chart row does not say which one it is. Thirteen titles in our Korean panel turned out to be foreign works. The test that found them is not the name — it is where they charted.

Across 801 matched pairs, a Korean series sits at rank 5.73 on the weeks a stablemate shares the chart and 5.2 on the weeks it is alone. The direction is the one schedulers fear. The size is half of one place out of ten.

Korean titles hold 7.9% of the world's Netflix top 10 places — 39,139 of 498,480. Of those, 82.3% carry the label we use to exclude same-name foreign works. 16.4% carry no label at all, and we keep them because we cannot check them.

Of 287 charting Korean series with a company credit, the 238 that aired on a Korean network reach a median 4 markets. The 49 with no broadcaster on record reach 39. They are 17.1% of the titles and 39.4% of the chart places.

A series charts in 32.9% of countries where its company had already charted and 8.6% where it had not. Comparing each series against itself removes age, strength and size, and the gap widens. It reverses in seven markets, five English.

Across 220 companies with a charting Korean title, catalogue size tracks total reach (r = 0.516) and not reach per title (r = −0.118). Firms with 20 or more titles get 11.1 markets from the median title; firms with two get 13.5.

Of 131 companies credited on the 345 Korean series that reached a Netflix country top 10, three account for half. Counting production credits alone — the answer to the obvious objection — it is 5 of 101, on a thinner base of 156 series.

In ten of thirteen years the two are within 20% of each other. Their year-on-year growth correlates at −0.13. Two industries arriving at the same number by unrelated routes, until 2024 pulled them apart.

Broadcasting holds staff 10.3 years, film 7.3, publishing and games 5.4. Pay runs the other way from what that suggests: film pays ₩63.1m against publishing's ₩77.4m while keeping people longer.

Platforms average ₩94.5bn in revenue; content providers average ₩1.5bn. Three-quarters of all webtoon businesses take less than ₩1bn a year, and 214 of them have fewer than ten staff.

Content firms average 23.9 years old against 44.3 for the rest of the listed market, and you cannot have long tenure at a young firm. Band companies by their own age and the tenure gap narrows from 1.86× to 1.32× — it never closes.

Korean titles drew between 74 and 79 million hours a week from Netflix's global chart in every year we hold. Exports rose 63% in the same two years.

Terrestrial broadcasters were 81.5% of Korea's TV exports in 2012 and 7.2% in 2024. Their staff average 10.3 years of service — the longest in Korean content. The producers who took the trade average 5.4 to 7.3.

396 webtoon businesses were surveyed in 2024. The 28 platforms average ₩94.5bn each; the 368 content providers average ₩1.5bn. The average describes almost nobody, and the studios that draw the comics sit at the bottom of it.