This year, I made two big decisions related to my music marketing agency:

1. I stopped offering press services.

I wrote about this here, but the upshot is that after five-plus years of helping artists get covered in blogs, I just got kind of burned out. I haven’t run a PR campaign in six or seven months, and I don’t miss it.

2. I started offering Meta ads management for Spotify promotion.

I’d been doing this for a handful of clients over the few years prior, but 2023 was the first year I started actively looking for new artists.

My roster is still pretty small, but I’m currently managing around 2-3k in monthly ad spend for clients, with the goal to get into the range of 10k per month in the near-ish future. You can read more about the service here (and apply to work with me, if you’re so inclined).

I’ve been very happy with both decisions.

Shoot, if all my choices worked out as well as these two, I’d be a GameStop millionaire and I wouldn’t have wasted ten hours watching Season 3 of Outer Banks. But you can’t win ’em all.

Anyway, for the purposes of this post, I’d like to share some of the things I’ve learned as a result of that second decision (to grow the Meta advertising side of my agency).

I’m definitely still learning this stuff. But I’ve run more ad campaigns for artists this year than I’d run in any year prior, and I’ve come away seeing things a little differently than I did seven or eight months ago.

In light of that, here are three hopefully-practical thoughts.

1. Simple stuff works.

I started out making Meta ads much more complicated than they needed to be.

I’d try three to five ad sets, all directed toward different interest targets. I’d run three to five different ads in each ad set. Taken together, this gave me anywhere from nine to 25 different combinations of ads / targets, many of which ran at an indie-level expenditure of around $10 per day.

Then I’d make things worse by aggressively optimizing (i.e., shutting down “poorly performing” ads to introduce new variants, changing my interest targets, etc.) after only a few days of data.

This made me feel kind of smart. It also made for a lot of work. But the problem was that it didn’t actually work all that well.

Here’s the thing: If you’re spending $10 per day, you really can’t know which ads and targets are working well after a couple of days. Or even a week, really. You’ll have assumptions, but you just won’t have enough data to draw significant conclusions. That’s especially true if you’re already stretching your budget across a bunch of different ads and ad sets.

So, over the past two or three months, I’ve been paring things back.

The details are a little technical, but the gist is that I’ve been applying Charley Tichenor’s 3:2:2 framework to my campaigns. It’s meant for e-commerce, but the principles work for music marketing, too

Generally, this approach means:

  • Running one or two campaigns at a time, each with only one ad set.
  • Running one dynamic ad inside of each campaign, using three videos, two primary headlines, and two primary texts.
  • Doing minimal interest targeting and allowing Facebook’s algorithm to find the best fits for your ads.

⬆️ That’s a very high-level synopsis. If you’re interested in drilling down, here’s a video where Charlie walks through the approach.

To be honest, I find the guy’s teaching style slightly obnoxious; something about his delivery makes me feel like he’s angry about how stupid I am.

(But also literally every single campaign I’ve run with his strategy has dominated my own attempts at optimization… so I guess the tone is justified.)

The takeaway: Don’t overcomplicate things. Sometimes the simple stuff works best.

2. The visuals are very, very important.

This has always been true. But when you’re using a broad-target approach like the 3:2:2 strategy, it’s even more true, because your ads are doing the heavy lifting for your targeting.

So they’ve got to be eye-catching – for the right eyes, especially.

I usually try three or four different things, depending on the assets that the artist has at their disposal:

  • Live performance videos
  • Music videos
  • Stock footage (but again, it has to be eye-catching)
  • Artist photos with motion effects

Pretty much all of the ad gurus I follow claim that music videos or live performance videos work best. In my own experience, I’ve found that is the case most of the time – but I’ve also run some of my all-time best campaigns with random stock footage or artist photos, too.

So I don’t think you necessarily need a video with your face in it. Again, I think the main thing is that whatever you go with has to a) catch eyes and b) appeal to the right audience.

And when I say the main thing, I mean the main thing. Like, in terms of impact on ad campaign performance, the visuals are probably more important than the actual quality of your song. Sad, but true.

(Although obviously if your song sucks, people aren’t going to listen to it ever again and all of your ad spend will be worthless.)

(You and I both know your song is great, though.)

3. Ads to playlists are (usually) cheaper.

I wrote a few weeks ago about my experiment with building a playlist network. It’s been a bit of a wild ride (I got attacked with bots!!), but overall it’s been a pretty solid success.

And I’ve had success running similar campaigns for clients, too.

I’m not entirely sure why it’s cheaper to get streams and followers on a playlist than it is to get streams and followers for your own music. But my hunch is that it has to do with a) how people consume music and b) the name recognition you can access with playlists.

Basically, when you run ads to a playlist, rather than asking people to stream a new track from an artist they’ve never heard of, you’re pitching them a vibe or a genre: “Songs for roadtrips,” or “Emo hip-hop,” or “Instrumental jazz.” Get your targeting right, and you’re showing your ads to people who are already into the sort of thing you’re offering.

Then you stack the list with big-name artists that fans of the genre will recognize, so that when people see the list they already know they’ll probably like the music.

As a result, both the ad click and the follow are easier sells.

The takeaway: I think every artist who runs Meta ads should consider building at least one genre-relevant playlist using this method.

And if you’re somebody in a pretty niche genre (i.e., experimental ambient music, or nu-jazz, or industrial post-rock), I think this should probably be your go-to Spotify strategy.

Okay, this post is getting too long.

I wanted to make this one practical; reading it back, I may have just made it confusing. Sorry.

Idk. Truthfully, this stuff is a little confusing. There’s a lot of jargon, and the Business Suite backend of Meta often makes me feel like a centenarian trying to figure out the whole computer thing. But I also want to emphasize that, given a couple of hours and access to YouTube, absolutely anyone can learn how to run Meta ads – and most artists should, because they really do work.

Relatedly, if you’ve got any questions on Meta (Facebook) ads, let me know. I still don’t consider myself an expert, but I do play one in this blog post, and I’d be happy to give you my two cents.

That’s it for this one. As always, here’s wishing you good luck.