Back in 2015, one of the more infamous of its sizable set of, er…um, erroneous metrics debacles, Facebook knowingly concealed that the calculation for Video consumption on its platform was misleading. In typical fancy-Facebook-naming fashion, Video consumption was dubbed “Average Duration of Video Viewed”, leaving marketers to innocently assume it meant ‘the total time spent watching a video divided by the total number of people who played the video’. Nowadays we know that assuming anything when it comes to Facebook is a no-no.
Not until June of 2016, after a year of advertiser complaints, did Facebook qualify the definition of the word “viewed” in the definition of “Average duration of video Viewed”. You see, a ‘view’ only counted if the video played for 3 or more seconds. So if 1000 people who watched a video on Facebook for :03+ for a total of 10,000 seconds, the AVV = :10. By Facebook’s estimation the inflated metric, which advertisers were relying on to allocate their precious marketing budgets, overstated actual viewership by between 60-80%. Ouch. Without the :03 rule in place, perhaps another 50 people, or 500 people, or 5000 people breezed right by that video without giving it much if any attention… and perhaps 6000 people in total saw it, for at least a split second / a total of 10,000 seconds and all of a sudden what was a :10 AVV rate was really = :1.6. Even the savviest of ad’smiths would be hard pressed to fit any site, sound, motion and emotion into the contours of a sliver of content 1.6 seconds long. After discovering the “discrepancy” Facebook fixed the “mistake” and changed the name of the metric to “Average Watch Time”. Since then it’s splintered into a slew of new video metrics.
We glossed over many things in the midst of digimani. We assumed that consumers didn’t want to be ‘interrupted’, but we’ve come to understand that actually depends on the medium, the message, the disposition of the consumer, and a host of other variables. We assumed that consumers wanted to have ‘conversations’ with brands but have come to learn that very few do. We assumed that traditional media was a blunt instrument and that precision, one to one messaging to the right person at the right place at the right time was something that constituted an ‘advancement’ in advertising. We now know it is certainly an advancement in advertising technology, but persuasion is just as much an art as it always has been. We assumed that online advertising, or any advertising for that matter, worked in insolation. We now know that they work in tandem. We even assumed, and probably still do, that insights into patterns of behavior can be fashioned into predictions. Maybe. Consumers are people and people are fickle, emotional, impulsive and Unpredictable. We are as different from one another as we are from ourselves given our situation and mood. Behavior can be influenced by advertising just as much, if not more, than advertising can be influenced by behavior.
So let’s stop making assumptions and start by asking ourselves, did we falsely assume the meaning of “duration” in the Average Duration of Video Viewed metric, as further defined by “seconds”? What is a second? To calculate it as 1/60th of 1/60th of 1/24th of 1 day would be inaccurate due to the irregularities in the speed of Earth’s spin caused by changes in the distribution of mater on its surface (like the tides for example). To be dead accurate, we’d need to consult the International Bureau of Weights and Measures which defines 1 second as “the duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the cesium 133 atom”.
What Facebook really ought to do is set the video metric gold standard by adopting Atomic Clocks. Yes! Call it VERCAO or Video Engagement Relative to Cesium Atom Ociliations….it doesn’t get any more precise (or fancier) than that.
