# Facebook reveals miscalculation in video-view count, showed advertisers inflated metrics for two whole years

Facebook has more than once reiterated that ‘Video’ is the future of the platform, and it is engineering models towards making that work. But, the social networking behemoth has come clean about a slight error in the video metrics it presented to advertisers on its platform.

The company has recently discovered a discrepency between the definition of average duration of video viewed and its calculation — resulting in an  artificial inflation in metrics for two solid years.

For those unaware, Facebook defines average duration of video viewed as,

“total time spent watching a video divided by the total number of people who have played the video.”

But, it had errornously calculated the average duration by dividing the total time spent watching your video by the number of views of your video for three or more seconds. And well, this is the wrong way to do it and that’s where the error lies. In this calculation, Facebook has ignored the millions(or billions!?) or users who viewed the video for under three seconds — thus causing a rift in average duration and average %age views.

Facebook revealed the error in an advertiser support web page, stating that the miscalculation has led to the delivery of an overestimated viewing time for over two years. Though it doesn’t mention the exact numbers that may have accounted for dragging the average down, but advertisers have told WSJ that it is likely to amount to about 60 to 80 percent. It has also mentioned the fact that the miscalculation, however, doesn’t affect your ad bilings in any way.

Average % of Video Viewed was also impacted since it used the Average Duration metric in its calculation so both metrics are being updated.

The social networking giant is now redifning the terms, and laying great emphasis on words — which they should do on calculation as well. The new metric defintions are as under:

• Video Average Watch Time is now being defined as the total watch time for your video, divided by the total number of video plays. This also includes the plays that start automatically and upon clicking by you.
• Video Percentage Watched reflects the percentage of your video somebody watches per session, averaged across all sessions of your video where the video auto-played or was clicked to play.

The corrected metrics are now available in your Ads Manager page, including data starting Aug 25th, 2016. Facebook also plans subject your previous video views through the new calculation lens, and update your legacy metrics in early October.

But the real question here is — Aren’t the advertisers concerned with the actual time for which people engaged with their videos? To know, if the numbers are highly bloated or not?

After this minor fiasco, one could definitely question the integrity of the video platform setup by Facebook but it is highly unlikely that the company was intentionally showing a bloated metric to promote their service or attract advertisers. But, the shocking fact about the whole scenario is that the company –working aggressively on video — let it continue for two full years and didn’t even come across the error sooner.

Also, there has been about a month since the discrepency was discovered and advertisers also seem to be pretty chill about everything. We don’t see anybody calling out Facebook on its fraud or suing the company for not showing the correct tally of views on the platform. This calm can, however, only be due to no losses in ad bilings but the trust is something that is easy to get lapsed — even with minor errors on this large a scale. So, Facebook now need to do some damage control and keep any such errors on the lay low for future reference.