How to Apply A/B Testing to Social Media: Your Go-to Guide

How to Apply A/B Testing to Social Media: Your Go-to Guide

Always be up to Date subscribe to updates - February 9, 2018

Wouldn’t it be great if we could test everything before fully committing to it?

Like when you go get ice cream and they let you sample a few flavors before deciding on the one you want. Or when you go to buy a new car and you can take a few test drives before landing on the one you want to take home.

Why?

Because it sucks picking an ice cream flavor that sounds good but tastes horrible. And it sucks more to buy a car that has terrible handling and poor visibility.

Comparing one ice cream flavor or test drive to another helps you make the best decision.

So it makes sense that when using social media for your business, you’d want to test what you’re publishing before building out an entire campaign or a content calendar full of updates.

Thankfully, there’s a specific process used to do this right: A/B testing.

What is social media A/B testing?

In general, A/B testing is simply a method of testing two items against one another: your “A” item (the control) and your “B” item (the variation). You then compare results and move forward with a selection of either A or B, depending on how well each performed.

In this type of testing, you’re only focusing on one change between A and B at a time, so you can determine the results of that specific change.

When applied to social media, A/B testing refers to running a control post A against a variable post B and determining how the change between the two affected performance.

Does the change to update B make a significant positive impact on performance over that of update A? Then it looks like you’ll be moving forward with update B’s format over update A’s when it comes to scheduling out your next social media campaign or content calendar.

First things first: Setting up your analytics

But before you get started testing, you’ll need to make sure you have an analytics platform set up to measure performance.

Some accounts like Twitter and Facebook offer built-in analytics you can organize based on your specific needs.

You can also use an analytics platform like Google Analytics to cover all of the accounts you’re using to A/B test.

Not an analytics expert?

No sweat.

There are tons of great resources out there than can guide you in setting up your accounts – and outline everything there is to know when it comes to reviewing and understanding your results.

Here are just a few that offer a good crash course:

Once you get your analytics in order, you’ll need to decide on what elements of your social media updates to change for your A/B tests and for what intended results.

A/B testing ideas

Thanks to the flexibility of social media, there’s plenty of room for creativity when it comes to A/B testing.

Here are a few testing ideas you can apply to your social updates, as well as the potential results of each, performance-wise.

1. With/without image

One idea for A/B testing applies to the use of images.

Images can affect tweet performance by inspiring higher levels of engagement.

To do this, you’d first share your control update (A) with text only. Then, you’d share the variable update (B) with an image and the same text used in update A.

Finally, measure and compare the level of engagement each update received.

2. Image vs. video

Another option would be to test whether a video earns higher levels of engagement than an image does.

For this test, you’d share your control update (A) with an image. Next, you’d share your variable update (B) with the same text, but replace the image with a video.

Measure and compare the level of engagement each earned.

3. Changes to the CTA

Interested in getting more click-throughs on your updates and more traffic to your website?

Making alterations to the call to action (CTA) between update A and B can help you determine which copy earns stronger results.

4. Changes to the hashtag(s)

Testing different hashtags between update A and B can affect the level of engagement between the two.

To do this, you’d share update A with one hashtag, then update B with a different hashtag.

Another option for this test would be to share update A without a hashtag and then use the same copy, but include a hashtag, when sharing update B.

5. Quote vs. blog title

Do you share updates linking back to articles or blog posts?

Try sharing update A with copy that includes a quote from the article + URL and then sharing update B with only the title of the article + URL to see which earns more shares and click-throughs.

6. Length of copy

Does your audience prefer updates that are longer or shorter?

You can share update A with a few longer sentences + URL and update B with just a short sentence of copy + URL and determine which receives more likes, shares, and click-throughs.

These are only a few ideas to get you started.

As you can see, social media offers you the freedom to try anything and everything when posting updates and then measure and evaluate performance.

Tips and best practices when A/B testing

One last thing…

There are a few best practices to keep in mind while performing your A/B tests. Following these guidelines will help you obtain the most accurate results:

  • Test more than one post.
    The best results are measured in terms of a trend vs. a single point in time. By testing a few updates with A characteristics and a few updates with B characteristics, you eliminate any random, one-off elements that may affect performance of a single tweet in time.
  • Keep post times consistent.
    Your audience may be more or less active at certain times of the day. Here’s how to find the best time to tweet. Posting update A at the same time of day as update B ensures your results will be impacted by the change you made, not the time of day you posted. If you want to avoid getting too specific on posting times, test groups of A posts and B posts all at random times. Chances are, you’ll be able to retrieve accurate results with the average of the results.
  • Stay organized.
    Clearly, if you’re performing multiple A/B tests comprised of multiple updates, things can get confusing. I recommend keeping a spreadsheet to save and track all variations of your updates, as well as your results. By documenting your A/B testing in this way, not only are you better able to organize the best performing posts of each test, you’re also creating a record of the tests you’ve tried. Then you won’t waste time repeating tests over and over.

Always be testing

Hopefully all of the above helps you realize not only the value of A/B testing, but also that it isn’t overly complicated or difficult to implement.

And that’s a good thing.

Because your A/B testing doesn’t end after one go around. No way.

As you know, the world of social media is constantly changing with the addition of new algorithms, platforms, post specs, etc.

So you’ve got to repeat and adjust your A/B testing over time to ensure your post performance keeps up with the rate of change.

But that’s the good news: If you aren’t a testing and analytics expert yet, pretty soon, you will be.
The bad news?

No more excuses for poorly performing social media updates in your campaigns and calendars.


Originally published February 09, 2018 - Updated March 19, 2018

SUBSCRIBE TO OUR NEWSLETTER

Signup to our monthly newsletter to get the best of our content with the latest
Conversion and A/B Testing resources right in your inbox.

  • 9 Feb, 2018
  • Posted by Chris Makara
  • 1 Tags
  • 0 Comments
Chris Makara

Written by Chris Makara

Since 2003, Chris Makara has developed a broad digital marketing background with a focus on SEO, Social Media, Automation and Analytics. He is the founder of Bulkly, a social media automation tool for individuals and small businesses. Feel free to find him on Google+,LinkedIn, or Twitter.

LOVE WHAT YOU JUST READ?

Have inspiration and insights you'd like to share with our tribe of testers? Convert is always on the look-out for awesome optimization content. Submit your pitch here and we will get back to you: https://goo.gl/forms/uqZFXQ1OW4X36tTb2

CATEGORIES Blogs

[hclightbox id='5' text='Anchor text']