April 6, 2017

How to A/B Test Your Content to Increase Engagement

analytics on laptop

First of all, what is A/B testing? A/B testing is the process of testing two web elements to establish which one performs better. It’s hard to establish exactly what your audience is looking for in your content and on your website and getting mixed advice from experts doesn’t make it any easier. That’s why looking at your audience itself – through A/B testing – is one of the best ways to establish what will increase your engagement.

Your website is doing well, viewers seem to like your content, but you could still use a little more engagement and conversions. Instead of guessing what your audience is looking for, checking your analytics, and then trying to figure out where the problem is, try A/B testing. You can use A/B testing to distribute variations of your content to your audience in order to establish which type of content they prefer.

Before you start A/B testing, there’s a few fundamental things you need to know:

  1. Know what results you’re aiming for. Are you testing for views, clicks, or shares? You have to know what you’re testing for beforehand in order to have accurate results. A video title that wins the most shares may be different than a video that wins that most click-throughs. Decide your goal first.
  2. Isolate one variable. Like a school science project, you can only test one variable at a time. Throw in a bunch of changes, and you’ll never be able to tell what actually impacted the results. For example, I want to test what call-to-action is the most effective at the end of our video marketing tips videos. In order to test this, everything but the CTA in the videos has to be exactly the same. Title, thumbnail image, content, length, etc. – it all has to be the same.
  3. Keep your test groups random. Make sure your groups have an even distribution of existing customers and potential customers. Because you’re pre-existing customers are more likely to engage with your content regardless of what you’re testing.

Because video is what we do, now we want to talk about ways you can A/B test your videos in order to maximize engagement with your viewers:

What you decided to test depends on the elements of your video. It also depends on where you feel there may be some issues. Maybe viewers don’t finish watching your videos? Maybe they never seem to follow your call-to-action? Maybe people aren’t even actually getting to your video in the first place?

Here are some ideas of video elements that are perfect for A/B testing:

  • Title
  • Call-to-action
  • Thumbnail image
  • Length
  • Placement of CTA/promotion
  • Promotional copy

You can basically test any element of your video as long as you’re able to do it randomly and keep all other elements the same. Here are some examples of why you’d want to test an element and how you could do it:

Testing the length of your video:

You have a really great video thats a little longer than usual, but thats because it’s full of important and interesting information. However, your viewers still seem to drop off after the start. Try and create a condensed version of the video that runs for half the amount of time. Let both videos run separately for at least and week and then compare their engagement rates.

Testing the thumbnail image:

After looking at your analytics, you see that people are seeing your video, but they just scroll by and never actually click play. Make sure you create something visually appealing as your thumbnail image that provides your viewers with a good idea of what the video is about. You can A/B test your thumbnails by using custom images and frozen frames of your video to find out which thumbnails earn the most clicks.

Testing your promotional copy:

You send out an email that has the link to a really useful video inside, but because your video is on a landing page, you have to convince your viewers to leave their email. This means you need to have a great subject line and email copy in order to compel people to stop what they’re doing, click your link, and play your video. This gets a little complicated to A/B test because you have to make sure you’re only changing one aspect of your promotional copy. Once you’ve decided what variable you want to change, you can use a mail service to send out variations of your email to a random group of your subscribers. The email that is most successful can then be sent to the rest of your list.


Once you’re done A/B testing, use your results to make any necessary changes. However, as marketers, your videos always have to be as effective as possible. So you’re never really done A/B testing. Once you’ve improved the variable you were testing, you can then move on to a different area that you think you might need to work on.


If you’re looking for more tips, check out our other blog posts here!


If you want to grow your business with video, call us at 321-802-3169 to set up a FREE video consultation.