With Digital video ad spend expected to grow 40 percent next year to $5.72 billion (eMarketer), Facebook’s sponsored autoplay video posts are widely touted as serious competition to Youtube for future digital video ad dollars.
So far Facebook’s video performance seems staggering, with clocked number of views on many videos outperforming the recorded views on Youtube. This is neatly summarised by Trevor Doerkson in his post Facebook TV its not about nothing.
“Another flick down the Facebook news feed five videos … were viewed 36 million times. The videos were all posted in the last four days & constitute a wild success. For comparison sake, here are some other big(ish) numbers:
- Super Bowl Sunday—as many as 118 million people will view the halftime show on TV
- Wednesday night—close to 12 million will watch Modern Family
- Thursday night—about 170,000 people will watch the LA Lakers on TV
- Saturday night—about 1 million people might view an epic UFC fight on TV
- YouTube – these same five videos received less than 10,000 views each
These reported numbers are incredibly interesting to those working in media who are compelled to deliver massive volumes of video views and reach on campaigns. But here’s the counter argument – this is impressions-based trading jargon, where huge numbers are used as a business driver and a measure of success.
Taking a closer look at the numbers on some campaigns we ran recently highlighted some very interesting results. Here are two screen shots from the promoted trailer using FB’s native player for the movie ‘It Follows’ on a budget of around 50GBP with fairly broad targeting.
What is interesting about this are the numbers:
- 30,000 people reached
- 6641 people watched the video – a staggering number when you consider the ad spend or 0.0078GBP per view
However when we take a closer look at these numbers 80% of users dropped off by the 5 second mark and only 7.48% (509) of people made it over the 30 second mark and 4.09% (279) of people watched the full trailer. Bearing in mind that you have to click the video to initiate sound there is a convincing argument that a large proportion left the trailer running with no sound and therefore were not watching it.
Things become more interesting when you look at the 3% (185) of people who ‘clicked to play’ (see the lighter grey line in the chart below) the video to initiate sound. Here we see a much higher level of engagement throughout the viewing experience where 67% watched more than 30 seconds and nearly 45% watched to the 2 minute mark.
- Behind the huge viewing numbers qualified viewing engagement is low proportionally to views unless there is a click
- This reflects how users use the news feed and browse information – trackings start the second the video starts video will be playing while users browse and read other content within their newsfeed
- Brands should therefore try and pack in the most exciting content in the first 10 seconds of the video play – this will change how video is produced for different platforms
Social media’s power in reaching and engaging audiences is unsurpassed in the realm of online marketing for entertainment content, and companies like Facebook are at the forefront for those agency advertising dollars. This strategy plays right into the agencies field of view that big numbers signify big glowing results, but behind the large numbers real engagement is actually far smaller and qualifying clicks lead to a much higher quality user. We are conducting more experiments in relation to what we are seeing – its tremendously interesting and I look forward to bringing your more data soon.