A/B Testing Best Practices with RadBot Video

You’ve spent the time and effort to create high quality content, but have you tested it? We recommend A/B testing for our video clients because it’s one of the simplest ways to ensure you are getting the best results from your content. A/B testing may seem time-consuming and daunting at first, but even a minor test can go a long way – generating more leads for your business. Here are some recommended best practices.

Plan Your List

There are lots of factors you can run an A/B test on. Some examples are subject line, header, color, layout, button size – the possibilities are nearly endless. In this example, we swapped out the actor playing the lead role, to see if the audience would respond better to the male or female lead. As you create your list, keep the factors simple and measurable. Prioritize your list by what you think will gain the most noticeable results, but don’t discount small changes. A simple color or size change can lead to noticeable results and takes little effort to execute. If you are working with a marketing or media company, share your list with them so that they can stay informed and provide feedback.

Perform One A/B Test at a Time

Once you have decided which A/B tests you are going to run, focus on testing only one factor at a time. This is important to know which change got you the desired results. If you are testing your call-to-action, don’t change your subject line at the same time. This will only confuse your results, making it difficult to know which change was more effective. You should also decide on the length of time you want to run each test. At a minimum, you should run an A/B test no fewer than 10 days long.

A/B Test on Multiple Platforms

Just like you wouldn’t limit your marketing plan to just one platform, A/B testing can be performed for just about any content piece on any platform. To make the most of your time and budget, start with the platforms that make the most sense for your industry. Also consider where you can automate the process, such as PPC (pay per click) ad platforms like Google Ads or Facebook. You can learn more about how to set up a Facebook A/B test here. Other great places to implement A/B testing is on your website, landing pages and email campaigns.

Don’t Make Assumptions Too Soon

When you start A/B testing, you’ll likely have a hypothesis for which option will win. If you begin to see the result you predicted, it may lead you to end the test too soon. If you have a very large audience, you may be able to perform an A/B test in a relatively short period of time. However, smaller data samples require more time.

There is also a randomness to A/B testing. Think of it as a coin toss – it takes a certain number of flips to see a trend emerge, but that trend could either reflect a preference or pure chance. This is why you should set a duration and/or audience reach goal at the beginning of your A/B test. Having clear goals in mind, you can avoid making assumptions about the results too early.

Make Changes Based on Results

As you work through multiple A/B tests, keep updating your content based on the results. When done correctly, you should start to see gradual improvement over time. If there are momentary decreases in your numbers, don’t get discouraged! It could that you aren’t testing for long enough, changes in the market or factors outside of your control. Remember that not all A/B tests will show results that are statistically significant, so focus your efforts on the changes that will have the greatest positive effect.

Are you regularly using A/B testing in your marketing strategy? Let us know in the comments!

-Sara E. Jones, CEO & Co-founder, RadBot Video

P.S. Need help with video marketing ideas for your brand? Download our free guide here.