A/B Test

How to A/B Test Social Copy for Better Results

There is absolutely a right and wrong way to write social posts. And just like any other content format, there are specific rules to follow in order to get the best results. You’ve likely heard about A/B testing.

In marketing, it’s the practice of comparing two similar messages that have one core difference to determine which garners a greater response from your audience.

A/B testing is an especially effective tool for social media because the audience is largely anonymous. Testing different messages and formats can help you home in on what your followers are most interested in and what drives them to engage with you.

There are several components to a social post that can be changed for an effective A/B test:

A/B Test: Format

There are a number of different formats you can test. Here are a few to get you started:

  1. Question vs. statement
  2. Positive (Always, Do) vs. Negative (Never, Don’t).
  3. CTA (click here) vs. Implied (link)
  4. Numeral (8 reasons why…) vs. spelled out (eight reasons why…)

A/B Test: Word Choice

While you always want write social posts in your brand voice, you still have some options when it comes to the words you choose. Try testing whether using industry lingo makes a difference in engagement levels.

A/B Test: Visuals

The most obvious options here are posts with and without a photo. But you can also test different styles of visuals. Try a photograph vs. a drawing, for example. Or images of products vs. images of people. You get the picture (see what I did there?).

A/B Test: Hashtags

You can test using hashtags against posts without, but I don’t recommend it. Hashtags can help you organize posts, and help a larger audience discover them. But I would recommend using 1-2 per post (and absolutely no more than three). If you have more than three in mind, create an A/B test using different hashtags in each.

A/B Test: Timing

While there are a number of online articles that claim to know the optimal post times for each social platform, the truth is that none of them are 100% accurate. There are so many variables that can impact the success of a social post, and every audience is different. That’s why nothing beats performing your own A/B tests – and time of day is no exception. We were getting a low response rate on a particular post, but one morning we noticed a much higher number of clicks. We realized that when scheduling our social posts, we’d inadvertently selected 10 PM instead of the more typical 10 AM. That fluke propelled the success of that content. We might never have known that our audience is engaged late in the evening otherwise. It pays to test!

A/B Test

Here’s a great example of an A/B test I ran for a client several months ago.

Both were written to support a blog post about a husband and wife franchise team who’d been interviewed about their success:

A/B Test

Both use the same photo, but “A” doesn’t use their names, and calls out the fact that they doubled their business in year one. “B” mentions them by first name and talks about their “top three secrets.” You can see from the results which one was the clear winner. I’d love to say it’s the photo, or the use of keywords like “secrets” and “profitable” but the truth is, it may just depend on your audience.

Using a social dashboard (we are big fans of SproutSocial) is a huge help here because you can schedule two different posts at the same time on the same day of the week to minimize other variables that might come into play and affect your results.

Social media platforms aren’t the only place to run A/B tests. You can use this tactic to determine how your audience responds to:

  • email marketing (Test various subject lines and Calls-to-Action)
  • banner ads (test headlines, CTAs, background colors and images)
  • landing pages (headlines, subheads, CTAs, visuals)

With the proliferation of real-time data, you can react in real time to determine the most successful messaging, and make edits on the fly to capture attention and drive traffic.

NOTE: When we began A/B testing social copy in Sprout Social, we discovered that if we shared a URL link in both A and B, the in-dashboard bit.ly link shortener applied the same shortened URL to both, so the number of clicks always ended up being the same number for A and B. The fine folks at Sprout do have a workaround though. Here’s how to do it:

Add an arbitrary hash to the end of your link (meaning, add “/?” to the end of your link followed by whatever word you’d like). For example, http://mybloglink.com becomes http://mybloglink.com/?testA. If the URL ends in an extension, don’t include the slash (i.e. .html?TestA).

For every blog or piece of content you write, we recommend writing a minimum of 10 social posts. Bump that up to 20, by writing two versions of each and running tests. Sprout Social comes in really handy here, because you can track different promotions with a simple tag, and easily check in to see how well posts for each are doing. Once you start seeing a pattern in the types of posts your audience prefers, you can quickly edit the content to incorporate that feedback and improve your results.

Want more information about A/B testing or writing effective social posts? We get pretty geeked up about all things social and marketing around here. Give me a shout at beth@secondstoryagency.com if you’d like to learn more about running your own A/B tests or writing effective copy.


About the author

Beth is a veteran marketing professional, award-winning writer, and lifelong learner. She gets pretty geeked up about the power of words, content, blogging, and social media. She is an amateur skier and a lover of craft beer and Chicago-style pizza. She loves her life and she loves her work, and she tries really hard to balance them.

Leave a Reply