Home » , , , ,

Facebook video metrics update will hit small publishers most


Share on Facebook0Tweet about this on TwitterShare on LinkedIn0Email this to someone

Facebook updated video metrics on page insights last week.  The changes  in terms of what page administrators and editors will now be able to see on the Insights tab:

  • Analyze metrics across a date range: Earlier, Facebook’s video insights did not let page administrators analyze metrics across a date range, compare video metrics against historical benchmarks or sort top videos by minutes viewed or views. The latest update makes up for that. In the past, video views for a specific time period only included videos that were uploaded within that time period. The latest update gives total views for a given time period (or a specific day within that time period), regardless of whether the video was uploaded in that window or at an earlier time.
  • Compare current video metrics against historical benchmarks
  • Aggregate for 3-second video views, instead of 10-second views. Note that Facebook counts a “view” as 3 seconds of video played. If there are N number of videos, Facebook discloses the total number of 3-second views across that all the videos that have been uploaded. The latest update also only includes aggregations for 3-second video views. Aggregate 10-second video views will now only be available in the API and at the individual video metric level. Page administrators will still be able to use Insights to see 10-second view counts but only for individual videos.
  • Sort top videos by minutes viewed or views: The latest update lets page administrators see their top videos based on the total number of minutes people spent watching each video received.

Why is Facebook doing this?

Advertisement

The attempt here seems to be to inflate the video view (impressions) count while suppressing video consumption data. Remember that Facebook has auto-play for videos and counts a view as 3 seconds, which makes it easier for Facebook to show higher view numbers compared to YouTube (which defines a view as 30-seconds of video play). Invariably, for the same number of instances of a user loading a video, a video on Facebook is more likely to show higher views than on YouTube. This is despite the fact that a 30-second parameter for measuring views is far more indicative of a view than 3 seconds when it comes to measuring video ad consumption.

The latest update to video metrics follows an update to video ad buying options. In February 2017, Facebook introduced three new ad buying options for advertisers:

  1. Completed-view buying: advertisers will only pay for video adverts that have been viewed in their entirety, for any duration up to ten seconds
  2. Two-second buying: where at least 50% of an advert’s pixels are in-view for two continuous seconds or longer
  3. Sound-on buying: advertisers will have the ability to buy sound-on video adverts

With no data available for 10-second views, smaller advertisers and publishers will now have to work with video impression data instead of actual consumption data.

Note that the changes to Video Insights come months after Facebook reported errors in counting time spent on videos, errors in measuring engagement on live videos, and errors in counting instant article views. Since then, publishing and advertising industry has demanded more transparency and third-party audits of Facebook’s metrics. Facebook on its part, has agreed to third-party audits.

Who does it impact?

Facebook video ads are 3 seconds and 2 seconds, compared to YouTube, which has 30-15-6 second ad slots as options. With every platform counting views differently, it has been difficult for advertisers to buy ads cross-platform. In such a fragmented publishing and advertising environment without any independent measures, most advertisers define a view as 10-second video play. Without access to this data, smaller publishers will be forced to either develop their own software to extract the data using Facebook API or pay to subscribe to social analytics tools. Either way, it puts the cost of standardizing and counting views on publishers.

Share on Facebook0Tweet about this on TwitterShare on LinkedIn0Email this to someone