Hope Horner, CEO and founder, Lemonlight Video Production, talks about how marketers can apply A/B testing techniques to ensure the digital videos they create and distribute are having the desired impact on the targeted audiences.
Marketing is nearly an entirely data-driven discipline, and for a good reason: The ability to collect data about the performance of particular campaigns or assets allows better optimization. In the never-ending quest to optimize, A/B testing is one of a marketer’s most valuable tools.
You might think of A/B testing in the context of email marketing or website design, but as video becomes more popular among marketersOpens a new window , it’s critical to understand how to test individual video elements, too.
Generally, everyone has their own unique tastes. You might love a video you created, but will your audience also enjoy itOpens a new window ? A/B testing helps you find out.
Learn More: Five Hot Video Marketing Tools Coming SoonOpens a new window
Test What You Don’t Know
A/B testing video often means making changes to the actual video content itself. These changes can be a little more intensive than shifting copy on a website or swapping out call-to-action buttons on an email template. That’s not to say it’s difficult; you’ll follow the same processes you would in a typical editing job.
Inevitably, there will be more variations to test in a video than you might find in other media, especially because videos often live within website pages or email campaigns. In those cases, you’ll test elements of your video and the variables within the web page or email you’d normally test. It’s not always easy to predict how changes will work together or how different combinations will affect results.
There are a number of essential video elements to get right, and small changes to each of them can have a huge impact on overall performance. Here are the most important elements to consider:
1. Length. In the age of 8-second attention spans, shorter is better. That said, you should plan on producing at least two videos every time you run a test.
Determining ideal length typically depends on the video in question. For instance, let’s say you’re filming an “About Us†video for a client’s website homepage. This type of video can be longer than other types of video content because it often conveys a company’s origin story. You might shoot a 90-second version that contains all of the powerful, compelling components you’re looking for, but you should also edit footage down to a 30-second version.
Why? The answer is simple: Although some viewers might be interested in a quick summary of your brand, others might be drawn to your story and values. You won’t know which version resonates with more people until you conduct A/B tests. To measure success, prioritize click-through rates. The version that amplifies your CTR is the one to run with.
2. Title. Titles are relatively easy to test and can have a big effect on overall video performance. You can test a title’s length, tone, word choice, capitalization, and any other element that might affect a viewer’s decision to watch.
However, be cautious about changing too many variables at once. Focus on isolating individual characteristics and determining how effective they are before testing other qualities.
3. Thumbnail. Because it’s a small, often overlooked video element, your thumbnail could actually have a major effect on viewership. Because thumbnails are so easy to change, we modify ours relatively often at Lemonlight; we use everything from a brand’s logo to a still shot or a product shot.
If you consider every video frame as a potential thumbnail candidate, you have a lot of options. Select five that might reflect different preferences or grab attention for various reasons. After you pick those thumbnail options out, test them over time.
Keep in mind that thumbnails often appear in different contexts: What works well on one platform might not work well on another. Viewers on YouTube might gravitate toward a shot of people talking, and Instagram users might prefer a brand’s logo. Don’t assume each platform’s audience is the same — they have different styles for a reason.
4. On-screen text. On-screen text is another element that’s fairly simple to change, and it can significantly enhance your video (alternatively, it could also be an annoyance to viewers). Before you test copy specifics, find out how your audience responds to a video without on-screen text. If it turns out that audience members prefer the text, edit copy elements during postproduction.
If you plan on testing copy, have a few different versions prepared in advance. It’ll be easier for your video editor to make all necessary changes at once instead of going back into a finished video and making multiple edits.
5. Music. If you plan on testing different tracks to go along with your video, alert your editor in advance. Like handling on-screen text, swapping out music is relatively easy in postproduction — unless, of course, the flow of your video is tied to the rhythm and beat of your music.
If your music simply supplies the audio ambience for on-screen action, you’ll have countless options to choose from. You might test songs from completely different genres or the same song played at varying volumes.
When it comes to audio elements, your results will be somewhat dependent on your distribution channels. Some platforms might automatically play your music when a viewer watches your video, but others might leave the sound muted unless the viewer turns it on. Perform A/B tests on a platform-by-platform basis.
6. Call to action. Your CTA is one of the most important elements of your video. Here are two ways to test it:
Edit Your Phrasing
First, consider a product demo video that’s intended to drive traffic to a page where prospects can purchase the product. You might include a “Buy Now†button at the end of your video, or you could try one that’s more specific to your product. For instance, if you’re selling high-end water bottles, your CTA might say: “Get a Water Bottle,†or “Get My Water Bottle.†The difference between both phrases is small, but even minor tweaks can have a big impact on performance.
Encourage Different Actions
You could also vary the actions you want people to take after watching your video to see which action leads to more conversions. Try a CTA that says: “Buy Now,†and then try a less direct approach, like: “View FAQ Page.†If the latter button gets more clicks, perhaps you should answer some common questions in your next video (or perhaps your product demo wasn’t quite clear enough).
Learn More: Automation vs Personalization: What SmartReply on YouTube Means for Content CreatorsOpens a new window
If another button — perhaps something like “Contact Us†— prevails, your viewers might want to speak with someone before they actually make a purchase. In any case, you’ll gain valuable insight into your sales process and a better understanding of how your customers decide whether to buy your product.
A/B testing provides a free glimpse into the minds of your audience members; you’ll learn exactly what their preferences are by seeing them vote with the actions they take. If you’re not running A/B tests on every video (and every other piece of content) you produce, you’re ignoring an opportunity to collect valuable insights. Don’t miss out on new and evolving engagementOpens a new window with prospects and customers — the knowledge you gain has the power to shape your future messaging and business.