People
Korean viewing is winner-take-all. Attention on the people in it is not.
One Netflix title is 21.3% of all Korean viewing. The most looked-up Korean actor is 3.1% of all attention. And within the fifty most-searched, appearing in more charting titles does not predict being looked up more.
We publish two counts that describe the same industry: how much Korean content the world watches, and how many people look up the performers in it. They do not have the same shape.
Viewing is a spike
Across 265 weeks of Netflix’s global chart, 236 Korean titles have drawn 23.7 billion hours. Squid Game alone accounts for 21.3% of that. The ten biggest titles account for 41.3%.
Attention is a slope
Now the people. Over 30 days we measured English Wikipedia pageviews for 1,008 Korean performers who appear in charting titles. They drew 7,463,168 views between them.
The most looked-up performer, So Ji-sub, drew 233,542 — 3.1% of the total.
| Share of all measured attention | |
|---|---|
| Top 1 | 3.1% |
| Top 3 | 5.9% |
| Top 5 | 8.5% |
| Top 10 | 13.8% |
| Top 20 | 21.7% |
| Top 50 | 38.8% |
One title is a fifth of all viewing. One person is a thirty-second of all attention. The fiftieth-ranked performer still draws a seventh of what the first does — where, among titles, the top one is larger than the next several combined.
And appearing in more of the hits does not help
The obvious hypothesis is that attention follows exposure: be in more of the shows the world is watching and more people will look you up.
Within the fifty most-viewed performers, the correlation between the number of charting titles they appear in and their pageviews is −0.16. That is no relationship at all, with a faint negative lean too small for a sample of fifty to mean anything by.
So the ensemble regular of four charting series and the lead of one are not separable on this measure. Whatever drives someone to type a name into Wikipedia, it is not simply how often that name has been on screen.
Why the difference is worth noticing
A title is a single object that everyone either watched or did not. Attention on a person is a thousand separate small decisions, spread across a month, and it does not collapse onto one winner the way a chart does.
That is a practical point for anyone reading Korean entertainment coverage. Rankings of titles and rankings of people look like the same kind of list and are not. A title chart is dominated by its top row and the rest is a long tail. A person chart is a gentle slope where the difference between first and fiftieth is a factor of seven.
One caution that applies to every figure above. 336 of the 1,344 performers on our roster have no English Wikipedia article, so they cannot be measured at all — they are not in the 1,008 and are not counted as zero. And a pageview count records that someone looked, never why. A month with one large piece of news about one person would redraw the top of this table entirely.
Where these numbers come from
Sources
- Wikimedia Foundation — Pageviews API — daily English Wikipedia article views, human traffic only, 2026-07-05 to 2026-08-03 · https://wikimedia.org/api/rest_v1/
- Wikidata — Cast member (P161) and country of citizenship (P27), used to build the roster of Korean performers appearing in charting titles · https://www.wikidata.org
- Netflix — Top 10 weekly global hours viewed (Tudum), used for the viewing-concentration comparison · https://www.netflix.com/tudum/top10
What we checked
- The roster is 1,344 Korean performers drawn from the cast lists of charting titles. 1,008 of them — 75.0% — have an English Wikipedia article that returns pageviews; the other 336 cannot be measured at all and are not counted as zero
- Those 1,008 drew 7,463,168 views across the 30 days. Every share quoted is against that total, so 'the top actor is 3.1%' means 3.1% of measured attention, not of all interest in Korean performers
- The viewing comparison uses our own published figures for the same catalogue: 236 Korean titles and 23.7 billion hours on Netflix's global chart, of which Squid Game is 21.3%. Those figures were corrected on 2026-08-07 and this article follows the corrected set
- The correlation between charting titles and pageviews is computed across the 50 most-viewed performers, which is the set for which we hold both fields. It is −0.16 — no relationship, with a slight negative lean that a sample of 50 cannot distinguish from none
What we left out, and why
- Anything about the 336 performers with no English Wikipedia article. They may be well known in Korea and simply unmeasurable here, and treating them as zero attention would be false
- Any reason for a view. A page climbs for an award, a casting, a marriage, a scandal or a death. This counts visits and cannot tell them apart, so no cause is attributed to any figure
- Career-long filmography. The title count is how many charting Netflix titles in our set a performer appears in — not how many things they have ever made. A veteran of forty films can show a count of one
- Korean-language and other-language Wikipedias. Only English is counted, which measures the English-reading audience specifically, and that is the audience this publication is for
- Any extrapolation from 30 days. This is one month, and a month containing a single large news event would look different
The data behind this
Written from the same data
- Korean film and television share actors freely — but less freely than chance alone would produce
- The panel is mostly television. The people who appear in it most are film actors.
- Two hops from Squid Game reaches 79% of every Korean title Netflix has charted
- Everything we can say about Korean casting rests on a median of five recorded names per title