How to Release a Song: The Independent Artist’s Complete Pre-Release Checklist

The timeline, backwards

Almost every avoidable problem in an independent release comes from the same cause: the work was done in the wrong order, or too late for the parts that need lead time. Distribution takes days to propagate. Playlist curators want the finished master weeks in advance. Registration bodies are slow. None of that can be compressed on the evening before.

A workable shape, counting backwards from release day:

• Week 8: the song is finished. Not nearly finished.
• Week 7: mixed and mastered, measured, faults fixed.
• Week 6: artwork done, title fixed, metadata written down once.
• Week 5: delivered to the distributor with the release date set.
• Week 4: registrations filed, pitch to Spotify editorial submitted, playlist pitching begins.
• Weeks 3 to 1: content shot and scheduled, pre-save live, countdown running.
• Release day and the two weeks after: the actual campaign.

If you have less time than that, do the stages in the same order and compress the gaps. What you must not do is reorder them, because each stage assumes the previous one is settled.

Stage 1: is the song finished?

The hardest stage to be honest about, because by this point you have heard the track hundreds of times and can no longer hear it as a stranger would.

Four things decide whether a song is finished, and each has its own test:

• The hook. Does anything stay with a listener after one play? Five tests for chorus strength, including the one that cannot be gamed: play it once for someone and ask them the next day what they remember.
• The words. Can you say what the song is about in one sentence, and does anything change between the first verse and the last? Six questions a critic asks of a lyric.
• The performance. Pitch, timing, diction, delivery and breath are five separate judgements, and each implies a different fix. How professionals judge a vocal take, and separately, how to tell whether the vocal is in tune without perfect pitch.
• The arrangement. Structure, length and contrast. Ten tips for a commercially competitive song.

Do not proceed until this stage is settled. Mastering cannot rescue a song with no hook, and promotion cannot rescue a song that nobody remembers. Every stage below is amplification: it makes whatever you have travel further, in both directions.

Stage 2: is the file ready?

This is the one stage with objective answers. Six values, computed from the audio rather than judged, decide whether the file will survive being encoded, normalised and played on a phone speaker.

The full pass is in how to know if your song is ready to release, including export settings and a final go or no-go check. In summary:

• Loudness around -14 LUFS integrated. What LUFS to target for Spotify, Apple Music and YouTube, and why mastering louder gains you nothing.
• True peak at or below -1 dBTP, and no clipped samples. How to spot clipping and true peak problems: one is permanent damage, the other only appears after the platform re-encodes.
• Loudness range above roughly 3 LU, so the track is not squashed flat.
• Phase correlation positive, for mono compatibility. And if the low mids are congested, the frequencies to cut.
• Spectral balance in line with references in your genre. If you cannot tell whether a problem belongs to the mix or the master, this is how to tell them apart.
• The vocal sits properly. If it disappears in the chorus, the cause is usually masking rather than level.

For the techniques behind the numbers rather than the checks themselves, ten mixing and mastering tips covers the wider ground.

Technical faults act as disqualifiers. A curator with a hundred submissions is looking for a reason to stop listening, and distortion in the first ten seconds is the easiest one available. The song never gets evaluated as a song.

Stage 3: artwork, title, metadata

Three things that feel like separate tasks and are really one decision made three times. They have to agree, and where they disagree you lose money and visibility in ways that are tedious to undo.

• The title. It is metadata as much as art: it decides whether people can find the song again after hearing it. Why a generic or malformed title works against you, including the formatting rules that get releases rejected.
• The artwork. It will be seen at fifty pixels far more often than at full size. How to derive a cover from the song and make it survive at thumbnail size, plus the delivery specs.
• The metadata string. Decide the exact title and artist name once, including capitalisation and any parenthetical, and write it down. Use it identically at the distributor, on video platforms, in artwork credits and on submission forms.

Concrete action: the single most avoidable technical loss for independent artists is the same song existing twice on a platform because a character differed somewhere. Streams and saves accumulate on separate entries, neither reaching the threshold that one would have.

Stage 4: who it is for

Positioning is the stage most often skipped, and skipping it makes every promotional decision arbitrary. If you cannot say who the song is for, you cannot tell which of two playlists is worth pitching, so you pitch both badly.

• The audience. A primary and a secondary group, plus the listening context where the song works best. How to define an audience that actually excludes people, and how to check the guess against your streaming dashboard.
• The comparable artists. Three names, sized correctly, on a stated dimension. How to build a comparable set that works in a pitch rather than one that flatters you.

These two feed directly into stage 6. A pitch that names artists already on a curator’s playlist is the fastest way to demonstrate fit, and it is impossible to write without having done this work first.

Stage 5: distribution and registration

Two separate things that get conflated, and the second is the one that costs artists real money by omission.

Distribution puts the recording on the platforms. Deliver a 24-bit WAV at your session sample rate, never an MP3, with the release date set far enough ahead that pitching is still possible. Most distributors want at least a week; Spotify’s editorial pitch requires a minimum of seven days before release and rewards more.

Registration is how you get paid for things your distributor does not handle. This is not new income, it is money already owed that goes uncollected because nobody filed. Seven ways one single earns beyond streaming royalties covers it in full, and the first three items are the urgent ones:

1. Join a collecting society as a songwriter and register the song with correct splits between co-writers.
2. Register for neighbouring rights as performer and as owner of the recording. A different body, a different pot.
3. Switch on content ID through your distributor for material you fully own.
4. Export an instrumental and a clean version while you still have the session open. A large share of sync placements need one or the other, on a deadline.

Stage 6: pitching and content

Everything up to here was preparation. This is the part most artists start with, which is why it so often fails: promotion amplifies whatever exists, and amplifying an unfinished record produces a wider audience for its faults.

Pitching runs on the longest lead time, so it goes first:

• Spotify editorial through Spotify for Artists, at least seven days out, with the similar-artists field filled in properly.
• Independent curators. Fourteen platforms for pitching to playlist curators, with costs and the reasons submissions get rejected.

Content runs on production time, so it goes second:

• Decide what to post per platform. TikTok, Instagram, X, YouTube and Spotify Canvas each reward a different format from the same source clip.
• Produce a week of it in one session. One song contains a dozen usable pieces before you invent anything.
• If there is a video, derive the concept from the song rather than borrowing one: the seven decisions in a video concept, and six formats you can shoot for nothing.

Concrete action: shoot the video, the vertical clips and the Spotify Canvas loop in the same session. The Canvas is eight seconds and it will be seen more times than anything else you make.

Stage 7: after release day

The most consequential mistake in independent releasing is treating release day as the finish line. It is the point at which platforms have finally gathered enough signal to work out who your song suits, and most artists go quiet exactly then.

• Be present on the day. Reply for a few hours. It is the one day where responsiveness genuinely changes the outcome.
• Post most heavily in the two weeks after. Re-cut the best performing clip with a different first second: each upload is evaluated independently.
• Watch saves rather than streams. A save is a commitment; a stream can be an accident. Your save ratio tells you what your audience actually wants next.
• Keep pitching. Independent curators accept released music, and the editorial window closing does not end the campaign.
• Then evaluate honestly. What is genuinely knowable about a track before release and what is not, so the next one starts from evidence rather than from a story about why this one did not land.

Releases are not events. They are campaigns with a long tail, and the artists who do consistently better are usually not writing better songs. They are the ones who did stages 1 to 5 before starting stage 6, and who kept going after the day everyone else stopped.

Check Your Track Before You Release