What Is Split Testing? A Beginners Guide For Musicians

It’s hard to know what your brand should look like, what your next single should be, what promo shots you should use. In fact, it’s just hard to know what to do in general.

I also feel like musicians and artists are worse at making these decisions than most people. Self-doubt and self-consciousness can cripple our sense of brand and our digital marketing decisions. At the same time, we tend to be stubborn and determined to go with our gut much of the time.

People – there is a better way. Marketers have been doing it for decades, and it’s called Split Testing. Split Testing is a method used to test different variables in ads or on a website to determine what performs best.

Note that split testing and A/B testing are the same thing. Some companies refer to it as split testing, and others as A/B testing. Many platforms that you already use, such as Facebook and MailChimp offer built-in split testing tools, but some of them just call it A/B testing. Do not be confused.

The variables you manipulate when split testing can be as simple as testing two different fonts on a landing page to see which one results in more conversions, or changing the color of a call to action link. The changes can be remarkable – sometimes resulting in twice as many clicks.

Split Testing As An Artist

In a way, artists split test things all the time. We show our friends and family a few demos that we think are good, and they let us know which ones they like best. Or we show them different promo shots, and let them decide which ones they like best.

The thing is, our friends and family are not always reliable judges of what actually works for a wider audience. They know your personal self, not the artist persona that you may want to project.

Instead, you should be split testing in front of your actual target demographic. For very little money, you can run two separate ads and see which one works best. You can test two different promo pics and actually see which one tests better.

It’s a cliché, but numbers don’t lie. If you run the same ad with two different promo pics, and one gets twice as many clicks, you know for sure which one works better. If you run the same ad with different text, you can start to figure out what “voice” works best.

Here are a few simple reasons why you should be testing:

• Split testing saves you money. By finding the most effective picture, text, placement, and audience, you get more clicks, impressions, and value for less. The worse your ad, the more it ends up costing you.
• Testing helps you make big decisions. Re-branding? Choosing an album name? Deciding on new photos? Find out what your fans actually like by split testing.
• Testing can help you keep your fans engaged. Ever wondered how to get more people to stay on your mailing list? How to get more people to respond to your posts? Testing can make everything you send and post more effective.

Split testing clearly has value for digital music marketing. Let’s get into the details.

Using Facebook For Split Testing

Facebook isn’t the only place you can split test, but it’s a great place to start. Recently, Facebook has made split testing an integral part of their Ad Manager. You can literally select the Split Testing option in the Ad Manager, and it will walk you through the process.

Your chosen Facebook audience gets split into random, non-overlapping audiences, who are then shown ad sets with one distinct difference, called a variable. The variable can be a different audience, placement, picture, text, etc.

You measure your results by setting campaign objectives, and then recording your results. Your objectives can be conversions, clicks, or traffic. When the test is over, you’ll receive a notification in Ads Manager or an email letting you know which ad won.

Currently, when you use the built-in Facebook split tester, you can only split test for audiences, delivery, and placements. This means you cannot split test different photos or text. This is annoying, but still useful. You can figure out how to maximize your budget and finesse your ad buys.

Just because you can’t use the split test tool to test different creative options, doesn’t mean you can’t split test creative options. Simply run two ads at the same time, targeted towards the exact same audience but with different creative.

That said, you should still try the split testing tool. The audience that you are showing your ad to, and the way it’s presented have a huge effect on how people engage with it. If you’re about to run a big campaign, you should have your audience figured out.

Your audience may change depending on the content of the ad and what you’re trying to accomplish. That’s why it’s important to run small test occasionally – you need to keep up with people’s changing tastes.

Facebook Split Testing Tips

• Run your ad for a minimum of three days. If you run the ad for under three days, the results will probably be insufficient to determine a winner. Facebook does not suggest running a split test for more than 14 days, as the results can usually be determined sooner than that.
• Don’t break the bank. Split tests are not your full-blown ad campaign. I would suggest paying no more than $5/day. $5 is enough to push the ad to your audience, but remains well within a budget. There is no minimum budget required for split testing.
• Before you make major decisions like re-branding or choosing new photos, try a small, low-budget split test. Even if it’s just for a few days with a budget of $10. Ultimately, you’ll end up with a product, brand, or ad that is better catered to your audience.

Split testing on Facebook will ultimately save you money and will help your ads and posts get more reach. You’ll also learn and develop a voice that works for your brand. It’s a win-win situation.

Split Testing On Your Website

Artists use their website for varying reasons and with varying regularity. However, if you’re using it for a download gate or to generate mailing list sign ups, it’s the perfect place to split test.

For example, if you’ve been trying to increase your Spotify followers, you could set up a landing page on your website that directs people to follow you on Spotify before getting access to a free download or merch item.

Instead of setting up one page, set up two pages with different wording or a different design. Do a limited ad campaign and test both versions before doing a larger campaign. This way, you’ll figure out how your money will be best used.

There are all kinds of creative ways to use your website to split test something, and you don't necessarily need to pay for ads. You could create different versions of your homepage and show one to 50% of your audience, and the other to the other 50%.

Split Testing To Find An Album Title

There’s all sorts of ways you can use the concept of split testing without actually spending any money on ad buys. One interesting concept that I came across, was the idea of using split testing to determine an album title by having your fans pick in a sneaky way.

If you were planning on using a song title as your album title, you should try this out:

Send an email to small group of fans. Maybe 40 or 50. Offer a free download of one song to each of them. Send them to a page with all of your songs listed. Then, you can track hits to each separate download page, and figure out which title appeals to your fans the most!

It’s a perfect blind test. Your fans don’t know which songs they’ll like better, so they’re basing their choice on name alone. You get to see which name jumped out at them first, and then you’ll end up with an album name that immediately appeals to the majority of your fans.

Split Testing With Your Mailing List

Mailing lists are still one of the best ways to get your music and your news directly into people’s line of sight. The problem is, many artists are terrible at sending email.

Sending email for marketing purposes is an art. Everybody has companies that send them promotions. You've probably noticed that that some companies are so annoying, you’re practically forced to unsubscribe.

Other companies (Waves Audio, for example) do a great job of their email marketing and actually keep me interested in their products. I’ve parted ways with several hundred dollars because of their emails.

When someone signs up for your mailing list, you want to keep them there, from the first welcome email through to full email campaigns.

Much like Facebook, most email list providers give you the option to split test. MailChimp has a built-in A/B tester that allows you to test all sorts of variables from content to templates to subject lines.

Although their tool works best with large lists of 20,000 subscribers or more, it can and should be used by those of us with humble 500 person lists.

Here are a few useful things to split test with your mailing list:

• Your welcome email. Try a few different subject lines, pictures and bits of content. Your goal is to prevent people from unsubscribing, and potentially direct them to other social media. Figure out a format that works for you.
• Test before big announcements. If you’re announcing a tour or a new album, you want as many people to see that as possible. Try out different headlines and formats.
• Test different frequencies of emails. You can design small segments of your list to be targeted. Send a small segment more frequent email than the rest of your list. Do they lose interest? Are you getting good responses?

Figuring out how to market yourself via email is hard, and many artists suffer from a lack of self-confidence when they’re doing it. Run your email marketing like an entrepreneur and you will see results.

Split Testing Makes Your Money Work Harder

I know it’s hard to put money behind ads and spend on marketing when money is in short supply. It can seem like a never-ending, vague effort. However, if you are interested in making your money work for you and spending it wisely, you should be testing.

You probably wouldn’t buy a $2,000 guitar without trying it first. So, why would you spend $100 on a Facebook ad that isn’t as effective as it could be. Or $1,200 on a re-brand that your audience won’t even like.

In the long run, testing allows you to spend less money, because your money will be used more effectively. It also helps you refine your branding, find your voice on social media, and improve your intuition as a digital marketer.

I find that once people start split testing, they begin to enjoy the process. It’s fascinating to figure out what your audience likes and very rewarding to deliver content and ads that work.

The key to split testing is to keep doing it. You’ll get better and better at writing ads, choosing photos and presenting your content, but there is always room to improve.

I was beginning to think I had a good handle on what my fans like to see, but I recently did a re-brand and ended up with something that was totally not our style. I didn’t test it, I didn’t even ask for any opinions, I just did it.

The new photos and profile pictures (that cost over $1,500) ended up getting way less engagement than some free promo pictures that were taken after a show. Who would have thought? If I hadn’t started testing these, I never would have known.

Since switching the branding away from my very expensive mistake, I’ve seen much improved results. I've been learning from my mistakes!