One release in 2021, nineteen in 2025: what the cadence actually did

August 25, 2026 · sauvachi

One release in 2021. Nothing in 2022. Three in 2023, two in 2024, nineteen in 2025, and eleven in the first seven months of 2026. As of 20 July 2026 the Spotify profile reads 141,386 monthly listeners. Those numbers are the whole post — everything else here is me being careful about what they do and do not prove.

"Does releasing more music grow you on Spotify?" is the most-asked independent-artist question and almost every answer online is somebody's vibe. I can't settle it either. What I can do is show five years of dates that are public and permanent, put one measured figure next to them, and be specific about the four things that could explain the gap other than volume.

The count, year by year

I counted these off the catalogue on this site rather than off any summary. Worth saying: a count taken from a single store can differ by a title or two, because not every release is on every service and some stores group multi-track singles differently. These are release events, not track counts — a three-track single with sped-up and slowed versions counts once.

The gap is the part people skip past

MURDER went out on 22 December 2021 and the next release was mamb on 3 August 2023. Nineteen months. If you want the honest version of this argument, you have to look at that window first, because it is the cleanest natural experiment in the whole set: a catalogue of one, no new release events, no new-music window, nothing for a recommendation system to act on.

I do not have an audited monthly-listener series going back to 2021 — Spotify does not keep a public history of that figure, and any chart I drew of it now would be reconstruction rather than data. So I'm not going to pretend I can show you the line. What I can tell you is that a nineteen-month gap on a one-track catalogue is the worst possible position to be in, and that 2023's three releases arrived in a six-week cluster (3 August to 14 September) followed by another eleven months of nothing before mont blanc. Restarting is not the same as running.

January 2025 and what changed

TWO AFFAIRS on 23 January 2025, then CHROME seven days later on 30 January, then LOADING on 13 February. That three-release opening is the point where the pattern breaks, and the rest of 2025 held it: sixteen more release events across the remaining ten months, an average of roughly one every nineteen days.

So: the dense year and the current listener figure are correlated in my data. I cannot prove causation, and neither can anyone else selling you a cadence strategy. Here are the four confounders I'd raise if I were arguing against myself.

  • Catalogue depth compounds independently of release rate. More songs in the system means more entry points, more autoplay tails, more chances a track lands in someone's rotation. That would grow slowly even if I stopped tomorrow.
  • Collaborations move audiences that volume alone doesn't. 4 BRICKS DOWN with Augxst, CHOICES with NST and MJ 4hunna, Wrong with Sadra, Can't Complain with Shameone, To Kojayi (Remix) with Youngka, Lessons with psydfx, THE SPLIT with NST. Each of those puts the name in front of a listener base that was never mine. Some of the 2025–26 growth is theirs, not the cadence's.
  • The songs might just be better. Unfalsifiable, uncomfortable, probably partly true. A year in which you finish nineteen things is also a year in which you get more repetitions at the craft.
  • Monthly listeners is structurally biased toward recency. It's a rolling 28-day count of unique listeners. An artist who releases every three weeks is permanently inside a fresh window; an artist who releases annually is measured, eleven months out of twelve, on residue. Some of the difference between 2024 and 2025 is a measurement artefact, not an audience.

The one mechanism I'd still bet on

Strip the speculation out and one thing survives: a release is the only event that reliably pushes you at people who already opted in. Release Radar exists to deliver new music to followers and to listeners with a history with the artist. That is documented behaviour, not a theory. Nineteen releases is nineteen deliveries. Two releases is two.

Everything downstream — editorial consideration, algorithmic pickup, the four-to-seven-day window where a track either holds its save rate or doesn't — needs a release to hang off. You cannot get a first-week signal without a first week. That's the argument for cadence, and it's narrower than "post more, grow more."

What I'd change about my own 2025

Three pairs went out on consecutive days: FAKE LOVE on 30 March and LIKE DBE on 31 March; Trance on 19 June and you would kill on 20 June; Ghost Town on 10 July and GO on 11 July. Two releases one day apart do not get two new-music windows. They get one window and split the attention inside it, and the second one usually loses. If I were running 2025 again the count would be the same and the spacing would be seven to fourteen days minimum, with nothing landing inside another release's first 48 hours.

The other correction: volume is not a substitute for the second half of the job. Listener growth and follower growth are different curves, and the gap between them is where a lot of "my numbers went up but nothing happened" comes from — which is why I've been tracking it separately in the Follower Gap Ledger rather than folding it into a cadence story.

What this is honestly worth to you

If you take one thing: cadence is a floor, not an engine. Nineteen bad releases would have produced nineteen wasted windows and a catalogue nobody wants to walk into. What the density did, at best, is make sure the work that was good enough had a delivery mechanism and a measurement window instead of sitting in a folder while the profile went stale for nineteen months.

The set of dates above is checkable against any store. The 141,386 figure is a single dated reading and will be wrong by the time you read it. I'd rather publish both with their limits attached than round them into a lesson neither of them supports.