For decades, whether a record was heard depended on radio programmers. That function has largely moved to playlist placement, and the mechanics are less transparent than radio ever was.
The three kinds of playlist
Worth distinguishing because they work differently.
Editorial playlists, curated by staff at the platform. These have enormous followings and placement on a major one substantially changes an artist's numbers.
Algorithmic playlists, generated per listener from behaviour. These reach more people in aggregate and are less visible, since no two are the same.
And user playlists, made by listeners, which range from a handful of followers to audiences rivalling editorial ones.
Each is filled by a different process and each requires a different approach.
How editorial placement actually happens
More conventional than people assume.
Platforms provide submission tools allowing artists and labels to pitch unreleased tracks to editorial teams, with a window before release.
Editors receive far more submissions than slots and make decisions on the usual mixture of taste, fit, timing and relationships.
Label relationships matter, as they always did with radio, and independent artists do get placed, less often and less prominently.
The pitching process itself is straightforward and a substantial number of artists never use it, which is the easiest available mistake to avoid.
How algorithmic placement works
The larger and less controllable half.
Systems infer similarity from listening behaviour — what else people who play this also play — and from audio characteristics and metadata.
The signals that appear to matter are early engagement, save rate, completion, and whether listeners add the track to their own playlists.
Which means the first days after release disproportionately determine the trajectory, because early signals feed the systems that determine subsequent exposure.
This is why release strategy has become so focused on the first week, and why artists direct their existing audience toward saving rather than merely streaming.
The payola question
Worth addressing since the parallel with radio is obvious.
Direct payment for editorial placement is against platform policy and, in the radio context, was regulated after well-documented scandals.
What does exist is promotional tools where an artist or label can accept a lower per-stream rate in exchange for algorithmic promotion, which is disclosed and has been criticised as a variant of the same thing.
There is also a substantial grey market of third parties promising placement on user playlists for payment, much of which is fraudulent, and some of which uses artificial streaming that can result in the artist's track being removed and payments withheld.
The practical advice everybody in the industry gives is to avoid paid placement services entirely, because the downside is severe and the upside is generally fictional.
What it means for how music is made
The structural consequence.
Playlists are frequently mood or activity based rather than genre based, which favours tracks that fit a consistent atmosphere and disadvantages anything that disrupts a sequence.
Which creates pressure toward music that sits comfortably alongside other music, and away from anything that demands attention or shifts abruptly.
Several producers have described this directly, and it is visible in the homogeneity of certain playlist categories.
What actually works for an artist
From what people who have done it say.
Pitch through the official tools, every release, well ahead of the date.
Drive early saves rather than plays, since saves are a stronger signal.
Build an audience somewhere you control, because playlist placement is not a relationship and it ends without notice.
And treat placement as a spike rather than a foundation. Artists who sustain careers do it on repeat listeners, live audiences and direct support, and the artists who briefly appeared on a large playlist and then vanished are a large population.
The pre-save mechanic
A practice that has become standard and is worth understanding.
Artists ask listeners to save a track before release, so that it appears in their library the moment it is available.
The purpose is to concentrate listening into the first hours, generating the early engagement signals that determine algorithmic distribution.
It works, and it means the visible chart position of a release reflects campaign organisation as much as reception.
What the money actually looks like
Worth stating plainly since it is frequently misunderstood.
Per-stream rates are small fractions of a currency unit, and they vary by country, by subscription tier and by the total pool being divided.
The payment goes to the rights holder rather than to the artist directly, and what reaches the artist depends entirely on their contract.
Which means that discussions about streaming rates being too low are frequently really discussions about how the money is split once it arrives, and the two questions get conflated in a way that suits some parties in the chain considerably better than others.
The metadata that decides for you
An unglamorous point with real consequences.
Genre tags, mood descriptors, credits and territory information all feed the systems that decide where a track is placed, and they are supplied by whoever distributes the release.
Errors here are common and consequential. A track tagged into the wrong genre reaches the wrong listeners, and correcting it after release is slow.
Which makes the distribution paperwork a genuine part of the release rather than an administrative afterthought, and it is the part independent artists most often rush.