Kelly Gentile Thu Nov 1, 2018

Video viewing behavior today is vastly different from the traditional viewing behavior of video delivered on television or desktop ten years ago. With the rise of mobile experiences and the age of endless content, consumers expect video to be engaging, efficient, and valuable all the way through. Platforms like Facebook and Instagram are pushing this even further by implementing video into their interfaces in new and exciting ways compared to anything that has been done before.

In order to form data-driven best practices for Facebook video ads, SundaySky partnered with a major online retailer to test five different video story architectures to learn if one approach balances message and entertainment more effectively than the others when it comes to capturing and retaining the attention of audiences on Facebook and Instagram. The video format and in-video messages were consistent across all five Facebook video ads; with the duration at 15 seconds, the aspect ratio at 9:16 (for optimal vertical mobile viewing), and the audio kept to only music (since mobile viewers tend to watch without their volume turned on) in order to maintain an airtight testing environment.

The first of the five storytelling styles was linear.

This traditional video story arc is the approach most video ad spots use, and acted as the baseline (or control) for this test. In this style the story opens with entertainment while the messaging builds as the video progresses. For example, these Facebook video ads showed the primary product message first, followed by a secondary message (such as “free shipping”), and closed with pure entertainment by showcasing models wearing the latest fashion before revealing a “Shop Now” call-to-action.

The next storytelling style was the pulse arc.

The Pulse style punctuates (or “pulses”) the message every few seconds throughout a consistent entertainment flow. The Pulse video story is primarily an entertainment stream of different models wearing the latest fashions, and the product message interjects every few seconds by showing different camera angles of the product. The pulsing of the product message acts as a reveal of the accompanying messaging. This retains the viewer’s attention by displaying a new message after each beat, such as “day time,” “night time,” and “everything in between.”

The third storytelling architecture was the burst style.

The Burst story front-loads the video ad with the core message and then continues the video solely with entertaining content to keep the viewer engaged. This storytelling arc makes it so that the most valuable information is given to the viewer immediately and then everything that follows is just meant to add extra value to that key message.

The fourth style was the boomerang architecture.

The Boomerang style uses repetition to create a hypnotic effect within the video; while not a common storytelling style, it has become a popular approach in user-generated content on Instagram. This storytelling style gives viewers time to ruminate on messages since they are pulled into watching something repeat itself multiple times.

The final storytelling style was the retrograde arc.

The Retrograde story brings the end of the baseline story to the beginning, so rather than opening with the core messages, this
story ends with the product and reveals the story in reverse. While the baseline Linear story arc built from the core product message and added the secondary messages to it, this Retrograde story revealed the secondary messages first and closed with the product message.

Facebook video ads storytelling styles visual

Which storytelling style works best for Facebook video ads?

This test sought to determine which video ad storytelling style resonates best for the highest possible viewer engagement that motivates consumers to take action. But what does ‘action’ even mean?

Peoples’ actions attributed to a Facebook video ad may include engagement, clicks or conversions — all simple actions for a consumer that result in very different value back to the brand. The wrinkle in this specific test is that the different story techniques drove very different types of viewer actions — and there is merit to each of the five creative storytelling treatments.

That said: the Linear arc had the highest percentage of people who watched to the 95% mark, possibly because the audience is more familiar with this classic storytelling style. Pulse was the story style with the second highest view-through and retention. By its nature, Pulse has a fast-paced, driving beat to the story while the product message is delivered as mini reveals “pulsed” throughout the video. These results make these storytelling approaches stand out as the winners in the competition to attract consumer attention, since length of video viewed is a quite direct indicator of attention being held.

Meanwhile, the Boomerang video story style had the lowest percentage of people that reached the 95% view mark, which could speak to the repetitive nature of the story in which the audience understood the core message before reaching the end of the video. However, Boomerang had the highest percentage rate of people taking action (PTA). Although the specific breakdown of these actions is undisclosed, this high PTA rate may reflect the popularity of the Boomerang effect, which is now quite common in the social space.

Finally, the Pulse video story had the strongest return on ad spend (ROAS), followed very closely by the Linear video story. The ROAS metric is the one typically used by this brand to measure the success of Facebook video ad programs, so it could be argued that these storytelling styles are the real winners, but what does that mean to someone not using ROAS as a measure of success? In the end, what the test truly revealed is that different storytelling styles are best suited for different goals, and if you want your Facebook video ads to perform best in different programs with unique goals then considering storytelling style a modular aspect of your videos is the only best practice there is.


Leave a Reply

*
*