How music ad campaigns actually work
Most artists who write to us have already tried. They put fifty dollars behind a post, watched the views climb, got a few hearts and maybe a comment — and nothing moved on the track. The conclusion is almost always the same: advertising doesn’t work for music.
It works. It’s that “boost this post” and “run a campaign” are two different things, and the distance between them is this entire page.
What you’re actually buying
You aren’t buying streams. You’re buying attention, and sending it somewhere while hoping something survives the trip.
The chain looks like this: the ad gets shown to a number of people, some stop to watch, some of those tap, some of those actually open Spotify, some of those stay past the first few seconds, and a very small share saves the track. You lose people at every step. That’s normal and it doesn’t get fixed.
The craft is entirely in reducing the losses at the right steps, not in buying more impressions. You can double the budget and get a worse result, if the problem was the creative or the audience.
The technical problem almost nobody mentions
If you read one section here, make it this one, because it’s why so many self-run tests end badly.
You can’t put a pixel on Spotify. It’s a domain you don’t control, so there’s no way to install tracking code. Which means Meta and TikTok have no way of knowing what happens after someone clicks. They see the click, then darkness.
The consequence is heavy. Ad algorithms are excellent at optimizing toward a goal — but only when they can see the goal. Here they can’t. So you can’t tell the platform “bring me streams.” You can only tell it “bring me clicks” or “bring me landing page views,” and hope whoever clicks then listens.
Two things follow, and they separate people who know what they’re doing from people who are experimenting.
First: the real result has to be read somewhere else, in Spotify for Artists, and cross-referenced by hand with what the ads manager shows you. Two dashboards that don’t talk to each other, joined manually by you.
Second: the professional route is to send traffic through an intermediate page you own, where a pixel does fit, and on to Spotify from there. You add a step and lose some people on it, but in exchange the platform can see a real signal again and optimize on it. It’s a trade, and knowing when it’s worth making is already half the job.
Anyone who doesn’t know this optimizes toward the wrong number and finds out after spending. It’s also why services promising guaranteed streams for pocket change aren’t buying advertising at all: no real campaign can guarantee a figure, and what that shortcut actually costs we’ve written up separately.
The pieces you need, in order
An ad account in good standing. It sounds obvious and it isn’t: new accounts start with spending limits, and music is a category where automated review runs more nervous than usual. Repeated rejections can lock the account before you’ve begun.
Creatives. A still cover isn’t enough. You need motion — a visualizer, a slice of video, something that stops the thumb in the first two seconds. And you need several, because one alone burns out fast.
The audience. This is where everything is decided. Too broad and you pay for people who’ll skip; too narrow and costs spike because you’re buying the same people every day.
Structure and budget. How many campaigns, how many ad sets, how much to each, and above all how long to leave them alone before touching anything.
Reading the numbers. Knowing which ones matter and which are noise. It’s the skill that takes longest to build.
Instagram, TikTok or YouTube: does it matter?
It matters a lot, and not the way most people expect.
There’s no best platform in the abstract. There’s the one where the audience for that genre lives, in that country, at that moment. A singer-songwriter and a club track don’t live in the same place, and what works in the US doesn’t automatically work in the UK or Canada.
The kind of attention you’re buying changes too. On TikTok people move fast and the video has to hold up on its own; on YouTube they arrive already willing to listen to something; Instagram and Facebook sit in between. The same ad, moved between platforms without being rethought, performs completely differently.
That’s why we don’t marry one platform: we keep them open and let the real cost per stream decide where it’s worth pushing, track by track.
Where the money burns when you start from zero
This isn’t a theoretical list. It’s what we most often hear from people who arrive after trying it alone.
The learning phase. Every new campaign starts blind and spends while it figures out who the responders resemble. Edit it during that phase and you restart it — and some people restart it daily, effectively paying for learning and never for results.
Random audiences. Dropping three famous artists into interests looks like the obvious move. It’s almost always the most expensive one: everyone is buying those interests, and the auction sets the price.
A single creative. Without alternatives you can’t tell whether the problem is the track or the ad. Meanwhile frequency climbs, the same people see the same thing six times, and cost per click doubles with no obvious explanation.
Turning it off too early, or too late. Both cost. The first throws away learning you already paid for; the second pours budget onto an audience that’s already saturated.
What it costs to learn
We’ll be honest even though it doesn’t help us: you can learn this. There’s nothing secret about it and no qualification required. What you need is budget to burn while learning, because the theory doesn’t come from watching videos — it comes from watching your own money disappear into tests that weren’t necessary.
A few hundred dollars go just to getting comfortable with the interface and the rejections. Then you need considerably more to reach a cost per stream that makes sense, because you get there by trial. And the part that stings: once you’ve finally understood your genre in your country, that’s what you’ve learned. Change genre or change market and much of the work starts again.
If your time and budget were infinite, it would be a sensible investment. Usually they aren’t, and that money was the money you wanted to spend on moving the music.
Three routes, plainly
Do it yourself. Makes sense if you have time, if you set aside budget explicitly for learning, and if the subject interests you on its own. Some artists get very good at it and never need anyone again. It’s a legitimate route.
A generalist agency. They know advertising, that part is real. But they’ve almost certainly never worked on traffic that can’t be tracked, and they’ll apply the ecommerce playbook to a case where it doesn’t fit. The bill arrives fast, because their rates are built for people selling products with margin.
Someone who only does this. Costs less than an agency and starts much further along, because the work of learning audiences and formats was already paid for by someone before you.
What “we’ve already done this” means in practice
We’ve been buying traffic for over twelve years, and for a good while now we’ve been doing it for music across several markets at once. In practice it comes down to four things.
The audiences are already tested, by genre and by country, with the real costs attached to each. We don’t start from a blank screen.
We know what to write. Which copy clears review, which promises get an ad rejected, which tone holds on TikTok and which on Facebook — not the same audience even when they’re the same people.
We know what working creatives look like. Formats, lengths, what has to happen in the first two seconds.
We know when to scale and when to stop. It’s the least visible part and the one that weighs most on the final result, because that’s where it’s decided whether the budget works or just gets consumed.
From there it’s a service, not a course: you pick the volume and we run it. Start with monthly listeners to your profile if the problem is the profile as a whole, or with album and playlist promotion if you have a full body of work to move. Streams usually start appearing within 24 to 72 hours, which is how long Spotify for Artists needs to update.
One thing we won’t do is promise a number. The figures on our plans are estimates drawn from past campaigns, not guarantees — for exactly the reason this whole page has been circling: real people decide, and nobody sells a guarantee on somebody else’s decision.
What advertising can’t do
Campaigns bring people. They don’t improve the song. If the track doesn’t hold whoever hears it, pushing it just lets more people discover it doesn’t land for them — and no budget on earth has ever saved a track that doesn’t hold.
When the track does hold, the problem is almost always just that not enough of the right people have heard it yet. That one you can solve by buying it.
The short version
Running campaigns yourself is possible, and for some people it’s the right call. It’s just that the price of entry isn’t a tool subscription: it’s the months and the thousands of dollars it takes to learn what someone else already learned, on a kind of traffic that refuses to be measured like everything else.
If you want us to look at your specific case — genre, market, what kind of release you have coming — get in touch. If we think you don’t need us, we’ll say so.