Start from one clip, not from ideas
The usual failure mode is treating each platform as a separate creative brief. You post to TikTok, then sit down to invent something different for Instagram, then run out of energy before X, and by day four the release has gone quiet.
Professionals work the other way round. One clip, adapted five ways. The creative decision is made once, at the start, and everything after it is formatting.
That decision is which 15 to 30 seconds of your song to use. It is not an aesthetic choice, it is a structural one, and it has a right answer: the segment where the song is most immediately compelling to somebody who has never heard it. In practice that is almost always the hook, occasionally a distinctive pre-chorus lift, and almost never the intro.
- Find the moment the track lifts and holds. Not a single hit, a section that sustains.
- Start slightly before it, by a beat or two, so the arrival happens inside the clip rather than at its edge.
- Prepare two or three variations of different lengths. Platforms behave differently and having options costs nothing once the source is chosen.
Concrete action: before making anything, write down the start and end timecode of your clip and the reason you chose it. Everything below assumes you have that.
TikTok
TikTok is the platform where the audio itself can travel, which makes it the highest-leverage surface for a musician and also the least forgiving.
- The first second decides everything. No title cards, no logo, no build. The hook is audible immediately or the viewer is gone.
- Vertical, full frame, captions burned in. A large share of viewing is sound-off until something earns the sound.
- Content angles that work: the making-of moment, the lyric that means something, the before and after of a production decision, the mistake you almost released, playing the part live.
- Post the same clip more than once with different openings. This is normal practice, not spam, and the algorithm treats each upload independently.
- Sound page matters. Make sure your track is correctly linked so anyone using it feeds the same sound rather than a duplicate.
Concrete action: make three versions of the same clip with three different first seconds: one starting on the vocal, one on the drop, one on a spoken line to camera. Post them across a week and keep the shape that performed.
The common mistake is treating Instagram as one place. It is three, and they serve different people.
Reels are the reach surface, and they behave much like TikTok: vertical, hook first, captions burned in. The same clip works with minimal adaptation, though reposting a video with a visible watermark from another platform is worth avoiding.
Stories are for the audience you already have, and their job is different: not discovery but participation. Polls, countdown stickers to release day, question boxes, the link to pre-save. Stories are also where the unpolished material belongs, because the format signals impermanence and the audience forgives more.
The feed is the permanent record. Cover art, release announcement, credits and thanks, the photograph that will still be there in a year. Low reach, high signal: this is what someone sees when they check whether you are a real artist.
Concrete action: use a countdown sticker in Stories starting three days out. It converts existing followers into release-day listeners better than any feed post, because it notifies them when the moment arrives.
X
X is the outlier in this list because it is the one platform where the text carries the post and the media supports it. Everywhere else the caption is secondary; here a good line with a mediocre clip outperforms a great clip with a lazy caption.
- Lead with a sentence worth reading on its own. A statement, a confession, a specific detail about the song.
- Keep the video short and put the link in a reply rather than the post, since posts with external links are distributed less.
- Threads work for process. The story of how the song was written, one step per post, with the clip at the end.
- Reply to people. On X the conversation is the distribution mechanism, more than the post itself.
Concrete action: write the caption first, before choosing the clip. If the sentence is not interesting without the video attached, it will not work with it either.
YouTube
YouTube covers two very different formats, and most independent artists use only the first.
Shorts take the same vertical clip and reach the same discovery-oriented audience as TikTok and Reels. There is no reason not to post there, and the cost is a re-upload.
The premiere is the underused one. Scheduling the full track or video as a premiere creates a moment: a countdown, a live chat, and everyone arriving at the same time rather than trickling in. That concentration matters because simultaneous engagement is a stronger signal than the same views spread over a week.
- Premiere time: pick the hour your existing audience is actually online, not the hour the internet is busiest.
- Countdown: a couple of minutes is enough. Long countdowns lose people.
- Be in the chat. The entire point is that you are there.
- Thumbnail and title carry the click. A face and three or four large words beat a cover art crop.
- First line of the description is the only part most people see. Put the hook there, not the credits.
Spotify Canvas
Canvas is the most consistently neglected free asset in music promotion: a 3 to 8 second looping vertical video that replaces the static cover art while your track plays inside the Spotify app.
It is worth doing for two reasons. It plays on every stream, which makes it the highest-frequency visual you own, and it makes the share screen a video rather than a still image, which noticeably increases the chance someone passes the song on.
- It loops, so design for a seamless join rather than a beginning and an end.
- No text and no logos. They break the loop illusion and Spotify discourages them.
- Keep the subject centred, since the interface overlays controls at the edges.
- It does not have to be filmed. A slow move across an object, a light shifting, an abstract texture that matches the mood all work.
Concrete action: shoot it at the same time as everything else. Eight seconds of usable footage takes ten minutes and it will be seen more times than any Reel you post.
Hashtags in three tiers
Hashtags are worth less than they used to be and are still worth using, provided you understand what each kind is doing.
- Primary: the tags that describe what the thing actually is. Your genre, your instrument, the format. These are how the platform classifies you, and classification matters more than reach.
- Trending: the large, currently active tags. Almost nobody discovers a small account through these, because the volume buries you within seconds. Use one or two, expect nothing.
- Niche: the small, specific tags where the total number of posts is measured in thousands rather than millions. This is the only tier where a new artist realistically gets found, because there is room to be seen.
Concrete action: build a list of five to eight niche tags that genuinely describe your corner of your genre, and reuse them. Chasing whatever is trending each week produces less than owning a small category consistently.
How often to post
Frequency questions get answered with numbers that ignore the only constraint that matters, which is whether you can sustain it. A schedule you abandon in week two is worse than a smaller one you keep.
A workable shape for a single release:
- Two to three weeks before: occasional process material. Not announcements, just evidence that something is being made.
- One week before: the pre-save goes live, the countdown starts in Stories, the first clip appears.
- Release day: everywhere at once, and be present to reply for a few hours. This is the one day where responsiveness genuinely matters.
- The two weeks after: the majority of your posts. Most artists stop on release day, which is precisely when the algorithm has finally worked out who to show the song to.
The last point is the one that changes outcomes. A release is not an event, it is a campaign, and the interesting part happens after the day everybody treats as the finish line.
Choosing the clip, the angles, the tags and the timing is a set of decisions rather than a matter of taste, and they can be prepared before the release rather than improvised during it. The structure of the track itself, including where the hook actually lands, is measurable from the audio, which means the highest-leverage choice in this whole process does not have to be a guess.