What Happens After a Song Breaks Out on Spotify? We Followed 1,461 Breakouts for Six Months

Six months after a release that doubled the artist's monthly listeners, the median artist was at 1.42 times their release-day number and had kept 19% of the gain. Only 32% of them kept half of it; 30% were below where they started. And the thing that decided which group you landed in wasn't how big the breakout was. It was how fast. Breakouts that peaked inside 30 days kept half the gain 19.5% of the time. Breakouts still climbing at days 60 to 90 kept it 49.5% of the time. Counted on 1,461 of the 3,863 breakouts from the breakout rate study, the ones old enough to have six months of history.

The reason I counted this one is the reading that comes after the good one. I've watched it with artists I work with: the single lands, the monthly listeners climb for a month, somebody screenshots it, and then there's the next month, and the number is going the other way, and the question is whether the whole thing was real. I've only ever had an opinion about that. Now I have a table.

The Data: Where 1,461 Breakouts Were Six Months Later

Every row below is a release that at least doubled the artist's Spotify monthly listeners within 90 days. Retention is how much of the gain was still there: the listeners at six months minus the release-day figure, divided by the peak minus the release-day figure. One means still at the peak. Zero means back where they started.

Six months after release	Breakouts (n 1,461)
Median listeners vs release day	1.42x
Same figure for releases that did not break out (n 18,889)	0.94x
Median retention of the gain	19%
Still above the 90-day peak	16%
Kept at least half the gain	32%
Kept less than half	39%
Below the release-day number	30%

At nine months, on the 438 breakouts old enough to measure, the median was 1.14 times release day, 26% had kept half, and 39% were below where they started. The release that didn't break out: 0.90 times.

Now the split that matters. Breakouts peak a median 40 days after release, with the middle half between 27 and 64 days. Here is what happened to each group:

When the 90-day peak landed	Breakouts	Share	Median peak vs release day	Median at six months vs release day	Kept half the gain	Still above the peak	Below release day
Inside 30 days	478	33%	2.81x	1.21x	19.5%	9.2%	37.0%
30 to 60 days	563	39%	3.02x	1.37x	28.4%	13.5%	31.4%
60 to 90 days	420	29%	3.04x	2.10x	49.5%	27.6%	19.5%

Same size of breakout, nearly, in every row. Completely different six months.

And the thing that didn't matter:

Size of the breakout	Breakouts	Kept half the gain at six months
Doubled (2 to 3x)	750	30.0%
3 to 5x	376	32.4%
5 to 10x	214	32.7%
More than 10x	121	36.4%

Those differences don't survive our holdout test. The timing result does: on 20 out of 20 random half-splits of the data, the early-peak group held its gain less often, and with the artist's starting size and genre held constant, an inside-30-days peak carried 61% lower odds of keeping half the gain.

By artist size, holding the gain ran 36% for artists under 2,000 listeners on release day, 31% to 32% from 2,000 to 50,000, 29% from 50,000 to 250,000, and 20% above that, though that last row is 55 artists.

What I Take From It

The median breakout is a reading, not a new address. Nineteen percent of the gain left after six months. The artist is still up on where they started, 1.42 times against 0.94 for everyone else, which is real and worth having. But if the number at the top of the curve got screenshotted as the new normal, the next six months are going to feel like a loss that isn't one. I'd rather you know the shape going in.

Spike versus build. A third of breakouts peak in the first month. Those are the moments: a playlist, a clip that caught, somebody with a following saying your name. Six months on, one in five of those artists has half the gain and more than a third are below where they began. The breakouts that were still climbing in month three are different animals. Half of them kept half, a quarter were still above their peak at six months, and the median artist in that group was sitting at more than double their release-day listeners half a year later. Same records, same yardstick, opposite outcomes.

Why, I can only read. My reading is that a spike is one source and a build is several. One playlist add brings people who came for the playlist; when the playlist moves on, so do they. A record that is still finding people in month three is being found through more than one door, and doors don't all close on the same day. It's the same thing I wrote about in why your song gets streams but no fans, seen from the other end.

Size of the climb is not the signal. I expected the tenfold breakouts to hold better. They hold about the same. If anything that's good news for the artist who doubled quietly: the quiet double that's still climbing at day 80 is a better position than the fifteenfold spike that peaked on day 12.

What I'd Do With This

If the number is climbing in month two, don't stop. That's the thing I see artists get wrong most: the campaign is planned around release week, and by the time the record is actually finding its people, the posts have stopped and the next single is being mixed. The breakouts that held were the ones that were still being found in month three. Give it the month.

If it spiked in week one and it's already coming down, don't panic and don't chase it. It was real, you keep some of it, and the next record starts from higher than the last one did. The 5,000 listeners study has the same shape at a different line: crossing and staying are two different problems. Plan the next release around the second one.

And if you want the next record to be the kind that gets found through more than one door, the rate calculator prices the mix and master to where you actually are as an artist, which after a breakout is usually a bigger number than you think.

How This Was Measured, and What It Doesn't Say

The data. The 51,868 releases from the breakout rate study, August 2025 through April 2026, each with the artist's Spotify monthly-listener history on both sides of the release date. The history runs to the end of August 2026, read on average every 11 days. 1,461 breakouts had a reading within three weeks of the six-month mark, which in practice means releases through February 2026; 438 had one at nine months. Twelve-month readings exist for 2 artists, so there is no twelve-month row.

The yardstick. Peak is the highest reading in the 90 days after release. Retention is (listeners at six months minus release-day listeners) divided by (peak minus release-day listeners). "Kept half" is retention of 0.5 or more. The timing claim was tested on twenty random half-splits of the breakouts, direction found on one half and checked on the other (20 of 20 agreed), and in a logistic model with starting size, genre lane and label as controls.

The edge of the window flatters the late group a little. A breakout still climbing on day 90 has its "peak" at the end of our window, so some of its later growth counts as holding the gain. The honest version of the result is spike against build, not day 60 against day 29, and the spike group's numbers don't depend on where the window ends.

Growth is measured at the artist, not the track. If the artist doubled in the 90 days after a release, that release gets the credit, and the retention is the artist's, whatever actually drove it.

The source can be wrong. Data is data, and sometimes the data can be wrong from the source. Chartmetric's pulling system doesn't always work, which is why I run my own systems next to it. A missed snapshot moves a baseline, and a moved baseline moves a retention number.

Correlation. Nothing here says that making a record build slowly makes the audience stay. It says the breakouts that built slowly in this window kept their audience far more often than the ones that spiked. I'm standing behind that sentence and no further.

Where the numbers come from: every figure above was counted from the study's per-release table joined to the artist history on 2026-10-08. The six- and nine-month tables, the peak-timing split, the breakout-size split, the holdout verdict and the adjusted odds ratio are in the study folder as study_09.json.