Posted: September 19, 2017
In 1984, if you wanted to make plans with your best friend, you had to pick up your see-through telephone and call her. If you were lucky, she’d have an answering machine you could leave a message on, or call waiting! Otherwise, you just had to wait and try again. Today, you probably text. Or, if you’re planning a big event and want to invite a lot of people, you use Facebook. Facebook is where people are – 72% of all adults use it. Even in healthcare marketing, if you want to reach a lot of people, using Facebook is a good bet.
So why do only 45% of marketers think their Facebook ads are effective? Understanding which ad element is working, and which isn’t can be tricky – there are just so many variables. But if you want to create more effective Facebook ads, you need insight into not only just what ads and elements are successful but why they’re successful. The best way to uncover this is through focused A/B testing.
A/B testing, also called split testing, helps marketers find the small details that make one creative item more successful than another. In order to effectively A/B test you should focus on testing one element at a time, in order. With Facebook ads, you have five elements that you can test:
• Headline Copy
• Call-to-action (CTA) button
• Link Copy
• Description Copy
Next, we’ll walk through some examples.
This is the most visible A/B test. Aside from the image, everything else is held constant in these two ads. Images can be a great method for testing demographics. If we want to see if a woman performs better than a man, if a younger woman performs better than an older woman – these can be tested with images, more so than age or gender-specific copy.
Testing Headline Copy
Here you can play with very small variations in verbiage or adjective use. Other times you might want to try a very serious headline against something less formal or even silly.
Testing CTA & Link Copy
Word choice can make or break your click-through rates. Facebook has a large number of CTA button options, but not all of them will be appropriate for your ad. If your link/CTA copy & button coordinate, the repetition will have a positive impact on your audience. Seeing words and phrases twice draws the eye.
Testing Description Copy
Testing description copy can be deceptive. It’s easy to slack on this copy or neglect to test it as frequently as images or headlines. Often, as users scroll through Facebook, they go right past the headline, but they might pause at the image and link copy combo. They might take a quick look through your description copy next. If they don’t like it or it isn’t relevant to them, they’ll keep scrolling. Choosing something interesting, informative, and short (a single sentence) could get you the clicks and conversions you want.
Other Factors to Consider
In the context of Facebook, frequency refers to the number of times a user scrolls past your ad. Think about Facebook ads like a billboard along the highway. The first time you drive by it, you might glance at it but not really take notice. The second time, you might read a line or two, perhaps looking at the image. The third or fourth time you pass by, you might be familiar with the idea of the ad and actually read through it enough to make a decision on whether or not you will follow its call to action. On Facebook, you should aim for a frequency of three or four when running your ads.
Facebook’s A/B Test
Facebook’s algorithm for A/B testing is proprietary and, as such, is a mystery to most marketers. What we do know is that Facebook decides very quickly which ad it thinks is winning. If an ad has been running for 24 hours and one variation received a few clicks, Facebook will decide that ad is performing better. It will throttle the impressions of the second ad, essentially creating a self-fulfilling prophecy. While this may seem harsh – not being able to give both ads equal real estate – you can still take a deep dive into the performance drivers other than frequency and spend. While frequency and spend will often be higher for one ad, the link click-through rate (LCTR), link cost-per-click (LCPC), and cost per result will really tell the true story of which ad is the winner.
The most important thing in A/B testing is to have a consistent measure of success that corresponds with the goals of the campaign. Depending on campaign type, and even the person performing the analytics, you could focus more on testing multiple images, different headlines, or any of the other copy to see what drives consumer interest. It’s a very subjective process. You know your company, you know what you’re trying to accomplish. If you’re testing images and there’s one that’s consistently performing better than anything else, but it’s been running for months and you’re sick of it, change it. Don’t get so hung up on the A/B test results that they become the be-all-end-all; they are an extremely useful tool for determining the potential interests and preferences of Facebook users.