I Take Back Everything I Said About Spotify Marquee

When Spotify launched Marquee two years ago, I was positively giddy. My first post on the subject proclaimed that it was even better than Meta ads.

Finally I had a way to reconnect with my listeners on Spotify, or at least those deemed part of my “reachable audience.”

That reachable audience quickly climbed from 2.1K to 4.4K listeners in the US. Despite the small audience size, I routinely managed to spend over $200 at $0.50 per click.

Today that same audience of previous listeners hovers around 23K, but Spotify no longer refers to them as my “reachable audience.” And for good reason! They’re no longer reachable.

But I’m getting ahead of myself…

In January of 2023, with 28 Marquee campaigns under my belt, my follow-up post declared the contest between Marquee and Meta ads a toss-up.

I continued to pair every new release with a Marquee campaign, aiming for at least 2.5 streams per listener and a 25% or greater intent rate.

Today there’s no contest. Meta ads win by a mile.

If you’ve read this far and don’t know what Marquee is, it’s this:

I describe the campaign type in excruciating detail here and share a monster spreadsheet of my results here.

When I saw Marquee had expanded it’s targeting options, I couldn’t wait to test it.

I’d been so busy with mixing and mastering work that I hadn’t written a new post in nearly two months. The combination of guilt and curiosity led me to splurge on a $500 campaign.

My results? Seriously disappointing. Why?

Spotify removed the ability to target my previous listeners.

When Marquee first launched, you could only target people who had already heard your music.

Today I have to target every Spotify user in the selected country. In other words, I have no targeting options at all.

Previously I was able to target an audience as small as 2.1K listeners. Now a custom audience of 90K in the US isn’t big enough.

Whichever combination of boxes I check, I’m met with “there aren’t enough listeners in your selection to meet campaign requirements.” I’m left with an audience size of 55.2 million.

55K previous listeners in Brazil and 33K in Mexico don’t cut it either, so I’m stuck with audiences of 18.6M and 15.1M respectively.

But hey, now I can customize the background color.

So it’s basically a wash, right?

That’s not to say previous listeners are excluded. Presumably Spotify tries to reach them. It just does a far crappier job of it than it used to.

Which brings me to my $500 campaign. Here’s the track if you’re in a contemplative mood:

And here are my results:

So much for splurging! Spotify only spent 10% of my $500 budget.

It’s not just me, either. My friend Dr Chrispy said he has a bunch of campaigns completed at $26, $42, etc.

Maybe, like with Discovery Mode, Marquee has gotten progressively worse due to high demand and limited supply.

My audience of 90K wasn’t big enough to meet their campaign requirements, but only 459 people saw the ad? Something doesn’t add up.

So roughly a 50/50 split between new and previous listeners, or in advertising parlance, cold and warm audiences.

87 out of the 459 people who saw the ad went on to stream the song.

What isn’t explicitly mentioned is that only 87 of the 148 people who clicked on the ad listened to the song, which means I paid for 61 worthless and likely accidental clicks at $0.35 a pop.

That brings the cost per listener to $0.60. Can I do better with Meta ads? Easily.

1.4 streams per listener with a 15% intent rate is far below the 2.5/25% intent rate benchmark I set in my previous post on Marquee. Keep in mind that users are being directed to my official This Is playlist (as opposed to my fake one), where even free listeners can easily stream more of my tracks.

On the other hand, the streams per listener for my other releases seems unbelievably high! 43 listeners x 19.35 streams per listener = 832 streams.

Contrast that to the 125 streams of the release I’m actually promoting.

To be fair, the numbers for “Disappear” aren’t drastically different than my previous campaign for a sonically similar single.

“Thoughts Can’t Hurt You” garnered 1.52 streams per listener with an intent rate of 18.75% reaching previous listeners only.

The difference? Spotify spent $202 of the $250 budget for 336 listeners.

One could argue that not spending my entire budget was a blessing in disguise, but I see value in reaching previous listeners beyond just the streams they generate.

Marquee, for me anyway, is now mostly indistinguishable from Showcase, except the latter will quickly blow through any budget you set (UPDATE: nope).

Should you bother with Marquee?

If you can still customize your audience, absolutely.

If not, you can always test a $100 budget on Marquee and/or Showcase, and compare that to your cost per conversion with Meta ads.

But it’s not an apples-to-apples comparison. With Meta ads, you’ll also get streams on other platforms, Shazams, followers, engagements on your posts, and the ability to actually retarget.

That said, some people never click on Facebook or Instagram ads, but will happily click on what appears to be a recommendation from Spotify from within the app.

Personally, I think I’ll reserve Marquee and Showcase for albums, where I always get the best results.

Are you using Marquee? Showcase? Meta Ads? Share your thoughts and strategies in the comments!