역주행: How a Korean Song Hits No.1 Two Years After Release

역주행 (yeokjuhaeng) literally means driving the wrong way down the road. In Korean music it means a song climbing the chart long after it was supposed to be finished.

Not a re-release. Not a remix. The same recording, unchanged, going to number one on its own two years later.

There is no equivalent word in English because the Anglophone charts largely cannot produce the phenomenon. Korea’s can, and does, often enough that the industry treats it as a real outcome rather than a fluke.

The case that defined 2026

“LOVE ATTACK” was released in August 2024, the title track of a rookie girl group’s first mini album. It did what most rookie releases do, which is not very much.

In July 2026 — roughly 680 days later — it went to number one on Melon’s TOP 100, Korea’s most-watched streaming chart. Korean outlets named the group the year’s defining reverse-run.

Here is what that looked like in search, which is the part you can see without a chart subscription.

Read the flat part first. For twenty-one months, from August 2024 through February 2026, the line sits between roughly 1 and 10 — small bumps at comeback releases, then straight back down. That is a normal rookie curve, and it is where most groups stay.

Then: March 2.6 → April 3.5 → May 37 → June 75 → July 100.

In four months it went from an eighth of IVE’s level to above both IVE and NewJeans. That is what a reverse-run looks like from outside.

The mechanism, in four parts

1. Korea’s main charts are realtime and streaming-weighted. Melon’s TOP 100 ranks what people are listening to now, not what was released recently. There is no release-window rule keeping a two-year-old track out. If a song is being streamed today, it competes today.

2. Short-form video is the ignition. In this case the trigger was traceable: a member opened a personal YouTube channel in February 2026, and clips built around a regional dialect and a specific styling persona spread rapidly. The chart followed the clips; the clips did not follow the chart.

3. Korean listening is concentrated. A relatively small number of streaming services, a relatively compact market, and a strong culture of listening to what is currently on the chart means attention compounds fast once it starts. The same concentration is why Korean fads also die fast — we have measured that shape before.

4. There is no gatekeeper to pass. Radio programming and label push matter less than they do in the US or UK. A song can arrive at number one without any industry actor having decided to put it there.

Why it lands differently from a debut hit

This is the part that is difficult to see from outside Korea, and it is the more interesting half.

A normal hit is delivered to the audience the label aimed at. A reverse-run arrives from everywhere at once, because short-form video does not respect demographic targeting. In the interview after the chart run, the group’s leader said the new fans ranged from primary-school children to people in their fifties.

Watch a Korean audience react to a group in this position and you can see three different things happening in the same room.

Older viewers feel 흐뭇함 (heutmutham) — a warm, slightly proud satisfaction, the feeling of watching someone’s persistence finally pay. There is no clean English word for it either; “gratifying” is close but too cool.

Younger viewers see the arithmetic of effort. Two years of promotion with almost nothing to show, and then the result. That reads as evidence rather than luck, and it is why reverse-runs generate unusually loyal fandoms.

Children see the style, and copy it. Which is how the styling and the dialect become memes independent of the song.

And underneath all three: the songs have to actually be good. A reverse-run cannot be manufactured, because the ignition is other people voluntarily making videos. If the track does not survive being heard for fifteen seconds by someone who has never heard of the group, nothing follows.

Other reverse-runs worth knowing

The pattern is not new, which is the point — this is a recognised route, not a one-off.

EXID, “Up & Down” (2014) — the canonical example. A fancam of one member performing to a nearly empty crowd spread online months after release, and the song went to number one. The group had been close to disbanding.

Brave Girls, “Rollin’” (2017 → 2021) — four years later. A YouTube compilation of the group performing at military bases went viral, and the song topped the charts in 2021. It is the longest gap most Koreans can name.

The shared shape: a small group, a song that did not get its hearing, and a piece of video that someone else made.

What this means if you follow K-pop from outside

The chart is not a release calendar. If you only watch new releases, you will miss the songs Korea is actually listening to. Melon’s realtime TOP 100 will regularly contain things that are years old.

A group being small is not evidence it will stay small. The twenty-one flat months in the chart above were not a failure signal. They were the normal condition of almost every group, right up until they were not.

When you discover a Korean group through a viral clip, check the release date of the song. You are often arriving at the reverse-run rather than at the release — which means there is a back catalogue waiting, and it means the Korean fandom is roughly as new as you are.

Related

• Is KPop Demon Hunters over in Korea? — the opposite curve, measured with the same data source
• What Koreans are actually searching for — how fast Korean fads rise and fall

What this article does not claim

Search interest is not chart position. The Naver DataLab chart shows how often people searched these group names. It is evidence that attention moved, and it lines up with the reported chart result, but it is not the chart itself. We do not have licensed access to Melon’s underlying data and have not reproduced it.

The index is relative. DataLab scales the highest point in a query to 100. All three groups were requested in one query, which is what makes them comparable to each other. These numbers cannot be converted into counts of searches, and they say nothing about listening outside Korea.

We chose the comparison groups. IVE and NewJeans are here because they are established acts with steady promotion, which makes their curves a useful baseline. A different pair would move the scale.

The causal story is reported, not measured. That a personal YouTube channel opened in February 2026 and that dialect-based short-form clips spread is what Korean media reported and what the group discussed in interviews. Our chart shows the timing is consistent with it. Consistency is not proof of cause.

The earlier examples are from memory of widely reported events, and we have not re-verified their chart dates against primary chart archives for this article. Treat EXID and Brave Girls as illustrations of the pattern rather than as sourced figures.

We are not covering endorsements, merchandise or product tie-ins. Those change constantly and stating them without an official announcement in hand would be exactly the kind of claim this site exists not to make.