Split Testing 101: How to Actually Split Test Facebook Ads

— November 11, 2018

We talk a lot about how important split testing is. Split testing (or “A/B testing”) allows you to test multiple variations of your ad campaigns to see your audience responds to, and from there you can start to work out why. Before you actually test your theories, after all, they are just that– theories.

But how exactly should you split test your campaigns? How many factors of your ads should you be testing, and how do you budget for it, and how intensive exactly do these tests need to be? In this post, we’re going to take a look at everything you need to know about split testing your Facebook Ad campaigns so you can get to the bottom of what works for your business and start maximizing your profit.

How Intensive Do Split Tests Really Need to Be?

When you’re creating your campaigns, it’s easy to want to test out a bunch of different ad campaigns, like throwing spaghetti at the wall to see what sticks. It’s why many marketers create as many distinct, unique campaigns as possible.

Unfortunately, the information this yields isn’t really accurate. Was it the pain point you appealed to in that one ad that caused the high conversion rate? Or was it the fact that you used a video, or that the copy was longer or a different style than it was on other ads? If you aren’t running intensive, detailed split tests from the beginning, all the other tests you’ll run won’t truly be accurate.

Because of this, you need to test everything possible about your ads. This includes:

  • Ad format
  • Visual components (including images, multiple images, and videos)
  • Factors like video length or number of images
  • Copy, including styles of copy (long vs. short, bulleted vs. paragraph), the messaging, offer, and appeals that you’re going for
  • Targeting, including retargeting and the strategies you’re using

You essentially want to be split testing all the factors that are going to determine the creatives and the audience you want to respond to it. And, most importantly, you want to test one ad aspect at a time.

These two ads would not be a strong split test. Everything about them is different, including the ad text, the headlines, the product, the features being advertised, and even the style of visual component.

Split Testing 101: How to Actually Split Test Facebook Ads

A better split test would look like this:

Split Testing 101: How to Actually Split Test Facebook Ads

How Exactly Do I Do This?

Facebook’s ad system is broken down into three parts: The Campaign, the Ad Set, and the Ad. Multiple ads can fall underneath the umbrella of a single Ad Set, and multiple Ad Sets can fall under a campaign. Campaigns define only the objectives, ad sets determine targeting, placement, budget, and schedule, and the ads dictate the creative.

In order to effectively split test your ads in a way that is easy to track, there’s a specific system that I follow. Other brands and marketers may have a different system, but here’s what I do:

  • I create a distinct campaign for each specific split test that I’m running. I label that campaign accordingly. It may read something like “Engagement campaign, testing copy.”
  • I’ll create at least four to six variations of a single ad, changing only a single element. If you’re testing the visual, change that up and leave the copy either the same or very similar.
  • If I’m testing copy, I create “sets” of ad copy that match in terms of pain points and emotional appeals, and write 4 headlines and 4-6 ad texts that can be mixed and matched fluently. This is a great way to test out styles of copy and specific features of your product or service to see what works.
  • If I’m testing audiences or placements, I’ll include different variations of ad content to see which each audience responds to.

How to Use Facebook’s Split Testing Tools

Facebook now has split testing tools that allow you to automatically create A/B tests to evaluate variables like creatives, targeting, placements, and more.

In order to use this feature instead of creating these tests manually the old fashioned way, you must enable the split testing option on the Campaign page.

Split Testing 101: How to Actually Split Test Facebook Ads

At the ad set level, you’ll choose which variable that you want to test. As a reminder, you can only test one variable per campaign.

Split Testing 101: How to Actually Split Test Facebook Ads

Once you do this, you’ll see the option to enter in variables for set “a” and “b,” with the option to create more versions for “c” and “d” and so on.

Split Testing 101: How to Actually Split Test Facebook Ads

How Should I Budget for Split Tests?

There is, of course, a downside to all these split tests: you still have to pay for the ads themselves, and now there’s a lot more of them.

The best thing to do is to create budgets for each individual campaign as if you were allocating funds to a single ad. In this case, you’re spending the funds not on a single ad or its theory, but all of its potential variations, too. Spend enough that you’ll be able to see actually results and learn something (typically around $ 150-200 per campaign is enough to get started). You’ll also want to give the campaigns enough time to even out, which will be at least a week and a half. Then take a look at the results, and adjust the campaigns if necessary.

Split Testing 101: How to Actually Split Test Facebook Ads

Note that Facebook’s Budget Optimization can’t be used at the same time as split tests. If you’re manually creating split tests instead of using their split testing tool, however, this feature will automatically allocate ad spend to ads that are performing best, automating the process a bit.

Are There Special Considerations When Split Testing Video Ads?

When you’re split testing videos, you’ll want to approach the tests the same way you would with any other content. Videos, however, have more components to keep in mind than other types of ads, which include:

  • Copy
  • Video length
  • Music vs. no background music, and styles of music
  • Video content
  • Visual elements like text overlay
  • Video style
  • Video dimensions (though mobile-oriented is always best)

Keep the headlines and ad text either very similar or consistent, and test multiple styles and lengths of video content.

Pay close attention to what audience niches respond to each types of video. Someone at the top of the funnel would be more interested in an introductory video, for example, while someone who has already purchased from you would rather see a video sharing what’s new.

Audiences can be finicky about videos, in particular, depending on where they are in the sales funnel, because the information that’s most relevant to them will be drastically different but also take center stage in video content. Keep that in mind, and pay particularly close attention to video targeting and retargeting.

Conclusion

Split testing is a necessary part of any and all marketing campaigns, including PPC content. And unfortunately, without in-depth, intensive, intentional split tests, you’ll be unable to really see which parts of the ad creative are working and what’s not.

Ultimately, split testing means that you’re going to need to have scalable ad campaigns from the beginning. You’ll want multiple sets of copy, multiple videos, multiple images to test. It’s important to take this into consideration when creating your campaigns, especially if you’re hiring specialized workers like copywriters, graphic designers, photographers, or videographers to create your content. Finding tools to help scale your video campaigns will be crucial, allowing you to dedicate more of your marketing budget to ad spend, where it will benefit you most.

What do you think? Do you use a similar process to split test your Facebook Ads? Which ad factors do you split test first? Share your thoughts in the comments below.

Digital & Social Articles on Business 2 Community

Author: Ana Gotter

View full profile ›

(55)