’Tis the season for reflection, and here at Wistia, we’re all about reflecting on past marketing campaigns (and sharing good tidings). In fact, we shared marketing lessons all throughout 2018, like how to solve common marketing challenges with video and our comprehensive guide to video marketing. So, when it came time to reflect on our biggest ad campaign to date, there was no shortage of lessons to learn from.
Last year we launched One, Ten, One-Hundred, our first-ever original series. We worked with Sandwich Video to produce three video ads to promote our video recorder with a production budget of $1K, $10K and $100K. And you may also recall that I shared tons of details around the ad strategy we put in place to support $111,000 worth of Soapbox ads. Even with a well-thought out advertising strategy, there was never a time when the campaign didn’t feel experimental. Questions like “What are the right metrics to evaluate performance?” and “Which ad formats best showcase the video ad and generate conversions?” kept me up at night.
After all, this type of campaign (and advertising budget!) was a first for Wistia, so the pressure was on to produce revenue-driving results. Luckily, having co-founders who support taking risks helped me view this campaign as a learning opportunity as well, so in the spirit of sharing, here are the top five lessons learned from promoting One, Ten, One-Hundred!
As a marketer, when I need inspiration for new ad campaigns strategies, I typically turn to Google or industry newsletters. Over the years I’ve found a lot of content about testing different elements of ads, including variations of copy and images, but I rarely find content about how other marketers are testing the video creative itself.
Testing the video creative itself might seem scary because it means you may need to create multiple video ads (which can be a deal-breaker for some folks!). But while time, resources, or budget can often hold marketers back from using video more often, luckily shooting high-quality videos has never been easier thanks to iPhones, simple production gear, and affordable editing software. Regardless of the resources or budget, there are a few things to keep in mind when testing video ads.
- Keep the unique selling proposition the same: Product benefits or features being promoted in your video need to be consistent across all video variations. How the message is delivered can vary, but what you are promising to deliver with your product, solution, or service needs to stay consistent across all video ad variations.
- Make sure that the video execution is different: If you have too few similarities in your video ads, you won’t know why one performed better over the other. That means you won’t get enough data to reach statistical significance to declare a winner. The best way to move the needle and get results is to have very different video executions.
- In the ad campaign, keep variables constant: The ad copy, targeting, and media budget, all need to be the same to ensure that you have clean test data–the only variable that should be different is the video creative itself.
For example, to test the Soapbox video ads, we ran all three videos (each with their own unique production quality levels) as single video ads, and kept the ad copy, targeting, and budget the same. In the end, we were able to see which of the three video ads resonated the best with our audiences — the $10K video.
Worried about leaving the lesser performing video ad on the cutting room floor? Don’t be. Even though the $1K and $100K video ads are no longer running, we’re still sharing results from the ads in blog posts just like this. My advice? Look for opportunities to repurpose your videos in other marketing efforts. Perhaps as a social media post or a product video on your homepage. You can even upload the video into Wistia’s A/B testing tool to test on your website with customizations like thumbnails, CTAs, and video length!
With Facebook Ads Manager and Google Adwords, you can get an ad campaign up and running quickly. But, when it comes to understanding how your campaigns are performing, it’s hard to know which metrics to look at. To complicate matters further, media partners report on metrics differently and without an understanding of what these metrics mean, you could be making decisions based on the wrong data.
“Media partners report on metrics differently, and without an understanding of what these metrics mean, you could be making decisions based on the wrong data.”
I encountered this very problem when pulling together results from the Soapbox video ad campaign. I had ads running on Facebook and YouTube, and at first glance, it looked like Facebook was generating far more video engagement and at a lower cost per engaged view. But, when I took a closer look at how each partner defined the metric, it changed the reporting.
Facebook Video Play: The number of times your video starts to play. This is counted for each impression of a video, and excludes replays.
When looking at the Soapbox video ad performance in Facebook, the video plays number was huge (over 900,000 video plays!). Based on Facebook’s definition, this metric was interesting, but not exactly an accurate measure for engagement. Instead of using the Video Plays as a KPI, I used “Video watches at 25%” as a better indicator. In the case of the Soapbox video ads, it helped me understand how efficiently I could reach engaged users; having watched 25% of the video ad or 30 seconds of the 2 minute Soapbox ads.
YouTube Video View: A Video View is counted when someone watches 30 seconds of your video (or the duration, if it’s shorter than 30 seconds) or interacts with your video, whichever comes first.
After looking at this definition and then again at the video ad performance in Adwords, it made sense that YouTube’s Video Views was only 10% of the Video Plays on Facebook — because it was an entirely different metric! I wanted to know how engaged viewers were with the video ad itself, so instead of using Video View as a KPI, I used “Video played to 25%” on YouTube instead, as it was closer to the “Video watches at 25%” on Facebook.
As I mentioned before in my recent ad strategy post, it can be tempting to compare Facebook and YouTube against each other. But, each media partner serves a different purpose in the buyer’s journey, and there can (and should be) a place in your media plan that includes both. Let’s take a look at why each partner is important.
“Each media partner serves a different purpose in the buyer’s journey, and there can (and should be) a place in your media plan that includes both.”
Given that most people log into Facebook to check their newsfeed and do a quick scan of their friend’s activities, this channel can be considered somewhat passive. In fact, the average watch time of the Soapbox ads on Facebook was 8 seconds vs 25 seconds on YouTube. But even though average watch time was lower than on YouTube, the value in running paid ads on Facebook comes in the form of impressions. Each impression provides a brand touchpoint in the buyer’s journey, and the more people that interact with our brand through various mediums, the more likely they are to purchase.
Here at Wistia, we see value in every interaction or impression that someone has with our brand; whether its seeing an ad, a social media post, meeting someone, or receiving an email. Being mindful of all these interactions, and creating consistency is an important part of our marketing strategy.
YouTube also drives brand awareness, but its role in the buyer’s journey is a little different than Facebook. YouTube is a destination for users who actively seek and consume video content. So it wasn’t surprising that YouTube had a longer average watch time than Facebook. The great thing about YouTube was that it provided the opportunity for Wistia to engage with users for longer.
The surprising part was that YouTube also had a nearly 73% lower cost per install than Facebook. Until this campaign, I had already written off YouTube as a channel for only big brands, but I learned that YouTube can generate conversion volume, and efficiently! Throughout the campaign, I viewed the advertising budget as channel agnostic, which means that budgets were not locked by channel and I could freely moved dollars around based on performance. For example, it was clear early on that YouTube was the most efficient channel, so after pausing underperforming Facebook ad formats, I moved dollars to fuel the ads on YouTube.
So, all that being said, while Facebook did have a higher cost per view and cost per install than YouTube, I understood its place in a buyer’s journey and it remained a part of the campaign!
At the start of the campaign, I established a primary KPI of “Cost per Soapbox install” and a secondary KPI of “Cost per 25% watched.” For cost per install, I wanted the media to achieve a $8 CPI, which was admittedly a bit arbitrary and based on other direct-response focused campaigns we’d run. Similarly, for “Cost per 25% watched,” I referenced previous video ad campaigns to come up with a range of $0.10 - $1.00.
After three weeks, it was obvious that it was going to be an uphill battle for our Facebook ads to reach an $8 CPI. So, I lowered the cost per install goal for Facebook to $30 (it was hovering at a $40 CPI so far) and looked for the CPI to trend in that direction. YouTube had a CPI of $11, so with some optimizations, a $8 cost per install goal seemed attainable. Similarly, for our second KPI it became clear that YouTube was tracking on the lower end of the range and Facebook on the higher side. Based on that intel, I readjusted the cost per 25% watched for YouTube to be $0.40 and for Facebook to be $1.00. I started learning fast that KPIs need to reflect the strength of each unique media partner, not a one-size fits all measure.
From all of the positive feedback we received from the video ads and behind the scenes coverage, it was clear that people were interested in what went into creating and promoting this kind of video series. And while the $10K video was the top performing ad when evaluated against our campaign KPIs, we also found that three ads in succession gave us the opportunity to share the bigger story behind the campaign.
When evaluating all the possible ad formats, I looked for those that provided us with the ability to include more context through ad copy and additional videos, beyond just one single video ad. I found that with two Facebook ad formats; carousel video ad and sequential video ads.
- Carousel video ad: All three video ads were displayed in the carousel ad format, starting with the $1K ad.
- Sequential video ad: If a user viewed the $1K video ad, then they were served the $10K video ad, and then the $100K video ad.
Having all three ads displayed in sequence gave us the chance to share the video ads as a series. Both of these ad formats were less efficient than the single video ad when compared to our KPIs, but there were other metrics, like impressions, that indicated that the video ad formats were resonating with our audience.
“Having all three ads displayed in sequence gave us the chance to share the video ads as a series.”
The moral of the story is, you may not see immediate impact from these types of ads, but look for other metrics such as time spent with brand (which can be calculated using metrics like YouTube’s Watch Time) to give you an indication of engagement with the brand. Plus, taking a risk with a new, more unconventional ad format means you can learn fast and iterate faster.
No matter the size of the ad or video production budget, these advertising lessons still remain true. And when it comes down to it, here at Wistia we view ad performance (whether good or bad) as feedback from our prospects or customers. This kind of feedback can help shape all parts of the business–from ad campaigns and content development, to promotional campaigns and even product updates.
Metrics aside, what really gets me most excited is the response that Wistia has received so far from the Soapbox video ads and the One, Ten, One Hundred original series. Given the positive reaction from our audience, we’re super excited to produce more content like One, Ten, One Hundred this year! Get ready for more content, and more lessons learned.