A/B Testing for Musicians: A Study in Facebook Ads

1055 15

You can’t trust your friends!

Musicians are a suspicious lot. We tend to hold fast to our gut instincts. We heed the counsel of close friends and die-hard fans. We can be defensive and protective of our artistic “purity.” We’re afraid to do anything that smacks of out-and-out marketing.

But sometimes you shouldn’t trust yourself. Sometimes you should ignore the advice of loved ones. Sometimes you just need to get scientific about things.

Trust the numbers! They don’t lie.

A/B testing allows artists to try out variable options in an advertising campaign so you know how to best direct your efforts.

You’ve probably already conducted a very simple version of  an A/B test when you decided what song should be your “single” to give away on your website, to put in the first slot of your audio-streaming player, etc. Chances are, you asked your band members, your family, your friends, and your fans which song out of a couple choices (selected by you) they liked best. But sometimes it is better to beware the feedback of family and friends.

The ones you trust can give you bad advice?

Sometimes the well-meaning, loving people in our lives just don’t think like marketers. Sometimes letting them add their two cents can derail your promo efforts. After all, they know you in an intimate way, flaws and all, and may not be comfortable with the idea that you’re trying to sell an image and persona to the world.

This ALMOST happened to Gabe Martin of the band Lou Leaves the Room, who explains his experience with A/B testing Facebook ads (and why it is so important):

I learned quite a bit with my small experiment.  First, your picture matters most.  When I created the page I wanted a “cool” picture.  So I put on headphones, made a pouty face, and shot a closeup with my phone.

That is my “cool shot.”

My wife saw the picture, was embarrased that I would even put up such a terrible picture and insisted on another shot. So she gave me a sweater, took me outside, and had an impromptu photo session.  Out came the more comtemplative, brooding shot.  That is my “serious shot.”

Then I decided to make an ad. I am trying to make a splash with my upcoming second album, so I decided to offer my first album for free to boost fans to my brand new Facebook page.  I created a small campaign, and made two ads.  Both had the same copy/text, but one had the “cool” picture and one had the “serious” picture.  Here is an example of the ad.


The ad with the “cool” shot was seen by twice as many people, clicked on twice as much and was half as expensive to run. It was very interesting. I suppose I need to take more candid cell phone shots for an effective ad campaign on Facebook.  Also, I was right and my wife was wrong.  She took it like a champ though.


The cost of running the ad was $30. It lead to nearly 100 fans, 200 plays, and 10-20 downloads. So, in my opinion it was money well spent.  A few more runs of the ad, and I might have a real site on my hands.

-Gabe Martin

The Moral of the Story

When you can run separate ads at the same time, you have the ability to see what photos, what text, what offers, what approach works best for you. Of the two (or more), choose the one that performs higher. But never stop tweaking! It is always a good idea to keep A/B testing, even when you settle on what seems like the best ad. They can always be better!

-Chris R. at CD Baby (with much help from Gabe Martin).

In this article

Join the Conversation

  • I love this. I'm becoming fascinated with "tests" like this. I just heard a great webcast on http://earn1k.com/ between Rahmit Seti and Tim Ferris. They go into great detail on how they used testing to do things like becoming a New York Times Best Seller. More testing articles!!!

  • Pingback: Tweets that mention A/B Testing for Musicians: A Study in Facebook Ads | DIY Musician -- Topsy.com()

  • Thanks CD Baby! I've been on the fence about the facebook ads. I didn't realize they could be so affordable.

  • Great article that really gets to the point with actionable advice!

    I've been experimenting with Facebook ads and find that the key metrics are click-through rate and conversion rate.

    Click-through rate is right there on the Facebook campaign page and shows you how often people click the ad. It's a much better metric that impressions (which facebook controls) or actual clicks (which is dependent on impressions).

    Conversion rate is something you have to calculate yourself: what percentage of people who clicked the ad actually performed the action you were intending (either "like"ing your page or giving you their email address or buying your song, etc.). That's why it's important to run each ad at separate times so you know which ad is the catalyst for the conversions at that particular time.

    Finally, I find the most successful ads (based on click-through and conversion) are ones that compare yourself to a well known artist. Target fans of that artist when you create the ad and then mention that artist in your description, like "Fans of Coldplay love Playcold! Download Playcold's newest CD!"

    But, for the most part, just have fun with it, experiment, and figure out what works for you!

  • Truthfully, I can't say how much I have learned. But I do know this,I wasn't overwhelmed by useless tidbits of nothingness, that is highlighted, in glitter "do you want to get rich like Me!" propaganda. What I did read seemed like it was straight forward and to the point!
    Thank You for speaking to me in the first person, and having a "practical put to use philosophy!"

    Charles Gray

  • Totally agree that testing is very important, even for the most marketing shy musicians.

    One test I like is to help choose album titles…

    You send an email to a small section of your fans and offer them a choice of a FREE track to download and then send them to a page with 10 "different" songs. Each song title should be a potential title for your album.

    As long as you make the free tune quite abstract you can give your fan the same free track whichever title that pick.

    But the magic is that if you track hits to each download page you have just gotten a totally un-biased market test as to which album title is going to be most appealing to your fans.

    Hope that kind of make sense 😉

    – Chris

  • Sigh

    Ya mean ya only gotta pay $30 to get 10 people to d'load a track? Ima gonna jump on that.

  • Pingback: Lazzia.com blog – Trust the numbers! They don’t lie.()

  • Great article Chris. It hits on the importance of monitoring and testing FB campaigns to get the most return for your investment. FB has a really great platform for artists who can target new fans by location, interests and on a CPC basis so you'll get branding impressions and only pay when someone clicks the add.

    The text in the ad is also another area you can test. Sample different descriptions or call to actions to see what gets the most clicks. Looking at your previous social media updates and mining through the text of updates that received a lot of feedback are great ways to get ideas for descriptions.

  • What a great article. You give me courage to take another step forward in the world of facebook marketing.

  • Question – why did his ad get twice as many impressions? I have used facebook ads a handful of times, and it seems to me that If he had the same bid on both ads and the same targets, he would get the same number of impressions.. unless facebook defaults to the first one when the two ads are tied for an impression? Anyone know?

    Scott Robinson
    Hotspur

  • I've found Jango a more cost-effective way of getting folks to listen to our music and become 'fans'. Jango fans are pretty 'lite' fans in my experience and we don't get to interact with them the way we can with FB fans (who actually sign up mostly 'cos they've seen a show and are therefore a bit more committed), but as a way to do a targetted internet blast before a tour, I've found Jango a good option to do the sort of thing that 'Lou leaves the Room' has tried to do.

  • We agree…This is an excellent article…We have found cool photos aren't the ones we thought they were.. Also with the ads…try not sell but rather promote the music or a give away works better than the "hey buy our music" angle.

  • Teak are starting to use the facebook ads and are pleased with the fnas we are collecting as a result. We regularly change them and each time it creats new interest

  • Pingback: DIY Agile Marketing for Indie Musicians | Solveig()