Facebook misled advertisers with faulty video metrics for over two years

Facebook has been using a broken video metric for over two years, that overestimated the consumption of video by users. This adversely affects video marketing campaigns on Facebook. Facebook introduced new metrics from August 25, that fixed the error in the data collection. Facebook claims the faulty metrics did not affect the billing, but advertisers could have invested disproportionately in Facebook videos based on the erroneous reporting of the performance of video. Advertisers and media houses with video marketing campaigns get a table that displays the performance of their videos across various metrics. These include Average Duration of Video Viewed, and Average Percent Video Viewed. Facebook based these metrics only on users who watched a video for three seconds or more. This meant that both the metrics were vastly inflated. A video could be viewed for more than 100 percent, if a viewer watched the same video multiple times. Now Facebook is introducing new metrics, called Video Average Watch Time and Video Percentage Watched, both of which will start counting all views, even those that are less than three seconds. The new metrics, with slightly changed names, will start doing what the original video metrics were supposed to do. Multiple views by a user still count for more than 100 percent of the video viewed. The changes were announced in a post in the Facebook Advertiser Help Center. According to a report in the Wall Street Journal, Publicis media, a major advertiser on Facebook, received a letter explaining the measuring methods used by Facebook. Facebook apparently overestimated the average video view time by 60 to 80 per cent. Facebook reporting its own video consumption metrics is unfair to advertisers. A more fair option is to open up the social network and provide tools for advertisers to externally measure the performance of their videos.