No country's Netflix chart is a signal for another's — 58 markets are standing in one queue

Somewhere in a distribution office there is a second monitor with a neighbouring country’s Netflix chart on it. The reasoning is old and reasonable: if a Korean series is climbing in Thailand this week, Vietnam is worth watching next week.

We can test that against 265 weeks of country charts. The answer is no, and the interesting part is why no.

Most of the time there is no “first”

Before asking who leads, we have to ask whether anybody does.

Take every occasion a Korean series reached the charts of two countries — 134,279 of them across 235 series. Compare the week it first appeared in each.

Netflix releases worldwide, and it shows. More than half the time a series lands on both charts in the same week, and there is nothing to lead. Whatever a leading indicator could be, it is confined to the remaining 42.4% from the outset.

Where one does come first, the direction is almost fixed

For the pairs where a gap exists, we measured how lopsided the ordering is. Zero means the two countries trade the lead evenly; 50 means one of them is always first. 946 pairs had at least 20 titles arriving in different weeks.

The median pair sits 42.3 points from even, out of a possible 50. When one of two countries takes a Korean series first, it is nearly always the same one of the two.

That looks, at first, like exactly the signal the second monitor was bought for.

It is not a signal. It is a queue.

Here is the test that separates the two.

If country charts genuinely pass signals to each other, the relationships should be local and sometimes circular: A leads B, B leads C, and C leads A, because each pair has its own reason.

If instead the markets simply differ in speed — some reach a top 10 within days of a release, others take weeks to accumulate enough viewing — then every pair is just two positions in one line, and loops should be impossible.

We checked every triple of countries where all three pairs have a direction. There are 8,593 of them across 58 markets.

8,591 are consistent. Two form a loop.

If A takes a title before B and B before C, then A takes it before C — essentially always. That is not a web of neighbours signalling to each other. That is 58 markets standing in a single line, in the same order, title after title.

So the answer to the scheduler’s question is no. A neighbour’s chart tells you nothing that your own position does not already tell you, and your own position does not change from one series to the next. You need one fact — am I early or late — and after that the second monitor is redundant.

What we are deliberately not publishing

We are not printing the order.

Ranking 58 markets by how quickly they take Korean titles would be a league table, and a league table is the one thing this publication does not make: it invites the reader to treat position as merit, when what is being measured here is a mixture of catalogue size, release patterns, chart volatility and how many people it takes to move a small country’s top 10. The finding is that an order exists and is stable. Which end a given market sits at belongs in that market’s own sheet, next to the reasons it might sit there.

What this cannot tell you

It cannot tell you why the order is what it is. “Arrival” here is a chart entry, not a release: Netflix does not publish release dates by country, so a market that appears late may have had the series available from day one and simply taken three weeks to accumulate enough viewing to place. A big country’s top 10 moves on a different number of viewers than a small one’s, and none of that is in these lists.

Two limits are worth naming. Of 4,278 country pairs that appear in the data at all, only 946 could be measured — a pair needs 20 titles that arrived in different weeks, and most pairs of countries have never shared that many Korean series with a gap between them. And the 20 is our threshold, not a property of the data.

Why we bothered

Because the second monitor costs attention, and attention spent watching a neighbour is attention not spent on the one question that actually matters for scheduling: where in the queue does my own market sit, and how long is the wait. That question has a stable answer. The neighbour does not.

The full tables — the same-week share, the lopsidedness distribution, the transitivity count and what had to be discarded: whether one chart leads another.