Most music ads don’t fail because the targeting was wrong. They fail because the click lands somewhere that can’t convert it, and nobody ever finds out which part broke. This is how to run a campaign where every click goes somewhere you can measure, and where the numbers afterwards tell you what to change.
Where the money actually leaks
An ad has to survive five steps before it turns into anything: someone sees it, someone taps it, the page loads, they choose a service, they press play. People drop out at every one of those steps, and never evenly. One campaign loses almost everybody at the creative. Another gets a healthy click-through rate and then loses the lot on the landing page.
If your ad points straight at a Spotify URL, you can see the first two steps in Ads Manager and the last one, vaguely, in Spotify for Artists a day later. Everything in the middle is invisible. That is why so many artists come out of a £50 test with nothing more useful than a feeling that ads don’t work for them.
Putting a page you control between the ad and the music closes that gap. Same click, same play, plus a number for every step in between — which is the difference between a test that teaches you something and one that just spends money.
Decide what one click is supposed to buy
Before you open Ads Manager, finish this sentence: this campaign is buying ___. There are only a few honest answers, and they need different setups.
Pre-saves, if the release is still ahead of you. The click buys a save that drops into someone’s library on release day. Playloudr pre-saves cover Spotify, and fans who opt in can be set to auto-save everything you put out afterwards.
Streams, if the song is already out. The click buys a listener choosing their own service on a smart link and pressing play.
Email addresses, if you would rather own an audience than keep renting one. The click buys a sign-up in exchange for something worth downloading, using a promo gate.
A campaign that tries to do all three ends up pointing at a page with too many choices on it, and produces numbers you can’t read. Pick one, run it, learn from it, then run the next one.
Send the traffic to a page you control
The single change that makes the biggest difference is not in Ads Manager at all. It is where the link goes.
A Playloudr landing page records the visit before the redirect, so you get visits, unique visitors, how many actually clicked through to a service, the split between services, the country, the device, and the channel the traffic came from. Your own visits are excluded, so testing your link doesn’t pollute the numbers. Read more on what to do with all that in music marketing analytics.
It also means your Meta Pixel and Conversions API can fire on a page you own. That is what lets Meta optimise towards people who actually click through rather than people who merely tap, and it is what gives you a retargeting audience for later.
To be clear about the division of labour: you build and run the campaign in Meta Ads Manager. Playloudr is the landing page, the tracking on it, and the reporting afterwards. If you want the longer version of how the two fit together, see Meta ads for musicians.
Setting the campaign up in Ads Manager
- Build the landing page first, and open it on your own phone. Tap it from inside Instagram, not just Safari — in-app browsers behave differently. Check the artwork, the title, and that every service button lands where it should.
- Connect your Meta Pixel and Conversions API to your Playloudr public pages. Without it, Meta is optimising blind and you have no retargeting audience when the campaign ends.
- One campaign, one objective. Traffic optimised for landing page views is a sensible starting point. Move to a conversion event only once the pixel has recorded enough of them to have something to learn from.
- Keep it to one or two ad sets. A small daily budget split across five ad sets never gathers enough data to get out of the learning phase, and you pay for that in wasted impressions.
- Put three or four creatives in the ad set and let Meta pick. Vary the first two seconds, the section of the song you use, and the caption. Do not vary the song.
- Use one URL, with UTM parameters on the end. Same link everywhere, tagged per source, so your link stats can separate ad traffic from your organic posts instead of lumping them together.
- Set an end date and an amount you are willing to lose. This is the cost of finding out, not an investment that has to pay back this week.
- Then leave it alone for three or four days. Editing an ad set restarts its learning, so daily fiddling keeps a campaign permanently in its most expensive phase.
Targeting that doesn’t burn budget
- Retarget before you prospect. People who already visited one of your pages are the cheapest audience you will ever have. Spend there first, then go looking for strangers.
- Choose countries with your own data, not ambition. Your link stats already show where your listeners are. Advertising into an expensive market where you have no listeners is the fastest way to spend £50 on nothing.
- Name artists a fan would genuinely list, not the biggest name in the genre. The stadium act’s interest audience is enormous, vague, and full of people who will never hear you out.
- Don’t stack ten interests in one ad set. If it works you won’t know why, and if it fails you won’t know what to remove.
- Build a lookalike only once you have a real source audience. A lookalike of forty people is noise wearing a technical name.
- Exclude yourself and your team. Playloudr already keeps your own visits out of your stats, but Meta will happily charge you for your own clicks.
Reading the numbers afterwards
| What you see | Where you see it | What to change |
|---|---|---|
| Plenty of impressions, almost no clicks | Ads Manager | The creative. Change the video or image before you touch the audience. |
| Clicks, but far fewer landing page views | Ads Manager | Something is dropping people on the way. Open the ad on your own phone from inside the app and watch what happens. |
| Visits, but few clicks through to a service | Playloudr link stats | The page itself — unclear artwork or title, or too many service buttons competing for one tap. |
| Service clicks, but streams stay flat | Link stats vs Spotify for Artists | An expectation mismatch. The ad is promising something the first fifteen seconds of the song doesn’t deliver. |
| One country working, the rest quiet | Country breakdown in link stats | Narrow the targeting to where it is working, and stop paying for the rest. |
| Good numbers, nothing after | Fans and timeline | Nothing about the ad. You need a next step on the page — an email capture or a pre-save — so the visit leaves something behind. |
The Ads Manager numbers and the landing page numbers answer different questions. You need both to work out which step is losing people.
Work out your own numbers instead of borrowing someone else’s
Every cost-per-click benchmark you read online was measured on a different song, in a different country, in a different quarter, against a different objective. Quoting it back at your own campaign tells you nothing. The only number worth managing to is yours.
Three divisions get you most of the way there. Spend divided by landing page views is what a visit costs you. Spend divided by the click-through count on your link is what a service click costs. Spend divided by sign-ups is what an email address costs.
Then decide, before the next campaign, what each of those is worth to you. An email address you can contact for years is plainly worth more than a single stream, so it can carry a higher cost. Write the ceiling down in advance; deciding afterwards is how campaigns get talked into looking successful.
Keep those three numbers for every campaign you run. After two campaigns you have a baseline. After four you have a benchmark that is actually about your music, which is more than any blog post can give you.
The usual ways clicks get wasted
Boosting a post instead of building a campaign. The boost button hands you almost no control over the objective and often optimises for engagement — you buy comments on a post rather than listeners for a song.
Pointing the ad straight at a streaming service. It feels like the shortest path, and it costs you every piece of information about what happened, plus the retargeting audience you would otherwise have kept.
Changing everything at once. New audience, new creative, new budget on the same day means the result is uninterpretable. Change one thing per round.
Advertising a song with nothing behind it. If a new listener arrives and there is no way to hear more, no way to follow, and no reason to leave an email address, you rented that person for three minutes. An artist front page gives the curious ones somewhere to go — see what to put on an artist landing page.
Letting the audience die with the budget. The campaign stops; the people it found don’t have to disappear. Anyone who left an email address is in your fan list with their activity attached, and can be synced to Klaviyo when you are ready to write to them. That is the part of the spend that keeps working next month.