Using A/B Testing To Inspire New Music Marketing Approaches

Ferren Gipson is a digital marketer specialising in the art and design sectors, writer for ArtCream, musician with Pandr Eyez and freelancer for those who want a more beautiful, usable online space.

I wear many hats. I do digital marketing, I’m a lifestyle blogger, and I also perform in a music duo. All of these roles seem very random (and probably are), but each one informs the other. I was just reading up on A/B testing, as I’m presently doing some homepage testing for a major retailer, and it occurred to me that this type of experiment holds relevancy for the music industry.

For those unfamiliar with the idea, A/B testing (or split testing), is a type of experiment in which you live test two or more strategies against each other to measure which performs better. A very basic online example might be that you want to test the color of a button to see which one gets more clicks. You would leave the other elements of the page normal, and test your control (or unaltered version) against your test variant (the altered version).

In this drawing, you can see a very basic illustration of our test versions, A and B. Visitors to the site would be randomly served one of these two pages, and after the pages have been shown equally to a selected sample size, we would theoretically have a clear “winner” of which button gets more clicks. From that information, you would be able to make an informed decision about which color button to use.

The idea of A/B testing in the music industry isn't completely foreign. Major labels are constantly developing new artists, trying to find the next big success story. Typically, artists at a similar stage in their career would be offered a similar level of support until one shows more promise; the decision is then made to invest more time/money into the artist that shows the most potential to generate good returns. Makes sense. Another way of phrasing that approach is, “throwing a lot of shit at the wall, and seeing what sticks.” It’s an expensive A/B test, but majors can afford it. What about independent artists and musicians, though?

I’m glad you asked.

The shit-throwing strategy applied by some majors isn't so relevant for independent labels, who often don’t have the resources to sign and develop numerous artists, and who also curate their artists based on different and varying criteria. Where A/B testing can come in handy is in getting serious about the changing state of the music industry. The press life-span of releases is getting shorter and shorter. You can go town finding think pieces about how quickly we’re consuming music if you like, but as a musician, I can already tell you it’s frustratingly fast. We’re talking days, and on any given website, a new release may only be on the homepage for a few hours.

That’s a problem, but the real problem is how we’re dealing with it. Presently, the industry strategy is to feed into it by expecting more music from artists at a quicker return. Every artist is different, but it can be a crippling expectation for many who like to spend time finessing their music before it hits masses. If you aren't quick enough, you run the risk of being left behind.

Independent artists are great hustlers for their music because they have to be. From a first-hand perspective, what I’d like to see is more experimentation about the way music is released. Even major label artists are experimenting in this area. I’m sure we all know about Beyonce’s secret album. Not only did no one know it was coming on that day, she also dropped the whole album at once. And if that wasn't enough, she released a video for every song! That’s a lot of variables to experiment with, and not every (read: any) artist has the pull, resources, and fan base of Beyonce, but there is a lesson to be learned here: there’s more than one way to do this.

The magic of A/B testing is that there may be a right answer, but that same answer won’t necessarily be correct for all cases. Every case needs it’s own test. As an independent artist, why not test release strategies? If you have 15 songs to work with, why not split them over three EPs and release them in three different ways to see what your fans react to best? The industry is changing all of the time, and there’s no one-size-fits-all strategy to propel an artist to stardom. Going forward, the most successful musicians are going to be the artists that are brave and flexible enough to experiment. Test everything and then go with what works. Music may be an artistic discipline, but that doesn't mean that you can make decisions based on data.