On September 22, 2016, Facebook revealed that its platform has miscalculated the average view time for videos served within its advertising platform. In a post written on Facebook’s Advertiser Help Center, the social giant revealed that the metric used for calculating the average time users spent watching videos was “artificially inflated” because it only factored in video views of more than three seconds.
How was the metric artificially inflated?
Because Facebook only counts a video view when a user watches for three seconds or longer, the Average Duration of Video Viewed metric did not include views of less than three seconds. Excluding shorter views inflates the averages to be higher than the actual average view duration.
Why are marketers and publishers upset about the metric inflation?
Some advertisers are upset because average video view time was used to evaluate creative performance and content, and in turn could have had an impact on spending on Facebook advertising over other video platforms like YouTube, Twitter, Snapchat, and even traditional TV. Facebook has strongly stated the miscalculation did not impact billing for advertisers – stating it is only an inaccuracy of measuring total video viewership and engagement; not related to inaccurately tracking how long people viewed a video (Facebook only charges for videos viewed for three seconds).
What has Facebook done to fix the inflation?
Facebook has replaced the Average Duration of Video Viewed metric with Average Watch Time, which factors in all video views (including those under three seconds). David Fischer, Facebook’s VP of business and marketing partnerships, stated: “We have also reviewed our other video metrics on the dashboard and have found that this has no impact on video numbers we have shared in the past, such as time spent watching video or the number of video views.”
WHAT THIS MEANS FOR MARKETERS
For marketers who have used the average view duration metric in reporting, it is important to caveat previous reports by noting that benchmarks of average view duration only factored in views longer than three seconds. Reports going forward should instead use the new Average Watch Time metric for evaluating video performance.
Also important to consider is that while average view duration can be a great metric to look at, after running a few campaigns there is less and less to learn from average duration of video viewed. Too many factors can affect view duration, including creative optimization, the total length of the video, or even whether a video is closed-captioned. Instead, prioritizing metrics like total video completions, impressions + CPM, and cost per view as primary KPIs will help prove value regardless of how average view duration is calculated.