Release strategy

How to promote a song, in the order that works

Most release plans are a list of tactics run in parallel and hoped over. Order beats list. Three of these steps only work if the one before them already happened, which is why running them simultaneously produces the same spend and a fraction of the result.

Before you promote anything

Two questions, and if either answer is no, promotion is premature.

Does the record hold past the hook? A clip can make anybody curious. A campaign only converts if the full track delivers what the eight seconds promised. If the hook is the only good part, no amount of distribution fixes it, and spending on it is the most expensive way to find out.

Is it cleared? Master and sync both, and on a video, the audiovisual work as well. Clearance is not a formality that happens alongside promotion. It is the thing that has to happen first, because campaigns get pulled on cleared-one-not-the-other more often than for any other reason.

The order, and why it is this order

1. Clear the record, which on the sync side means the publisher rather than the distributor. Weeks one to three before release, because a publisher who needs to come back takes time you cannot compress. Set clearance up against the catalogue rather than the single, so release two costs hours instead of weeks.
2. Map the hooks. Listen to the whole thing and timestamp the seconds that can carry a video alone. This is the step that decides the campaign, and it happens before a single asset is made.
3. Start short-form seven to ten days early. Not on release day. You want creations already pooling on the official sound page and follows already landing on the artist profile when the record drops, so release day lands into momentum rather than trying to manufacture it in twenty-four hours.
4. Release. The record goes live into an audience that has already encountered the hook somewhere.
5. Sustain for three to four weeks. A single burst produces a spike that resolves in a week and teaches recommendation systems nothing. Sustained posting produces a slope, which is what they are built to notice.
6. Pitch playlists once there is evidence. A curator seeing a record already moving on short-form is a different conversation from a cold submission. Pitching first and hoping is the standard mistake.
7. Read the data and reseed. Whichever hook held attention gets the volume. The others get dropped and their slots reallocated.

Steps three and six are the ones people reverse, and reversing them costs the campaign most of its value. The five-stage detail of step three is on how a clipping campaign works, and step six is covered on playlist pitching.

Where a small budget goes furthest

Say there is only enough for one thing. It is not ads, and it is not a playlist package either, which is unfortunate because those are the two things almost everybody spends the first money on.

It is getting a small number of real creator accounts, in the right genre lane, posting clips attached to one official sound inside a tight window. That produces a sound page with creations on it, which is the only asset in this entire list that keeps working after the spend stops. Ads stop the day the budget does. A sound page with four hundred creations on it is still there next month.

Ads work. This is an argument about order, and about what you still own at the end once the budget has stopped.

What to skip

• Anything guaranteeing a stream count. Spotify treats bought streams as artificial streaming and removes them. Not deliverable by anyone. Streams follow from a sound page being used, which makes them an outcome rather than a line item.
• Anything guaranteeing a playlist placement. Either it is a paid list nobody listens to, or the seller does not control the placement and is selling you a hope.
• Bought followers or bought views. They produce no completion, no saves and no follows, which means the algorithm reads a large audience that rejected the record. Worse than doing nothing. Verified views covers why.
• A folder of clips with nobody to post them. Distribution is the product. The edits are the input.
• Posting the same video from your own account fifteen times. One account cannot buy reach that a network can.

What to measure, week by week

Four numbers, in this order of usefulness.

• Sound creations. How many separate videos used your official sound. Moves first, usually inside days.
• Save rate. A save is a decision, a view is an accident of the scroll. This is the number that predicts next month.
• Profile follows. The one that compounds, because a follow puts your next release into people's feeds automatically.
• Verified views. Useful for cost accounting and the least predictive of the four.

Streaming movement comes later, weeks rather than days, and only after enough clips are in feeds. Anyone showing you streaming movement seventy-two hours in is showing you something else. How the Spotify algorithm reacts to short-form sets out exactly where the chain from a clip to a stream breaks.

Questions