Facebook, on Friday, said it has rectified an error in the way it calculated average duration of video viewed, one of the metrics available to marketers to gauge the efficacy of their video campaigns. The Wall Street Journal reports that this miscalculation was going on for two years.
According to Facebook, it first noticed this error about a month ago. The social media giant said that instead of the metric reflecting total time spent watching a video divided by the total number of people who played the video, it reflected the total time spent watching a video divided by only the number of “views” of a video. Facebook classifies a "view" as when a video is seen for 3 seconds or more.
This discrepancy could potentially have led marketers to think that Facebook videos are performing better than competitors as views will necessarily be a subset of total number of video plays.
However, Facebook says that it has already fixed the error last month and informed partners while also updating its video metrics. It further stated that this miscalculation "has not and will not going forward have an impact on billing or how media mix models value their Facebook video investments."
In a post on the Facebook Blog, David Fischer, VP (Business and Marketing Partnerships) at Facebook wrote, "While this is only one of the many metrics marketers look at, we take any mistake seriously. As soon as we discovered the discrepancy, we fixed it. We informed our partners and made sure to put a notice in the product itself so that anyone who went into their dashboard could understand our error. 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."
In a separate post in Facebook advertiser help center a month back, the company has further clarified the definition of its metrics. According to this post, both Average Duration of Video Viewed and the Average per cent of Video Viewed will be replaced with Video Average Watch Time (the total watch time for a video, divided by the total number of video plays, including autoplay) and Video Percentage Watched (percentage of video somebody watches per session, averaged across all sessions of the video) respectively.
So how much does it actually impact marketers and their ad spends? As reported by TechCrunch, Azania Andrews, Senior Director (Digital Connections) at AB Inbev, in a tweet about the controversy, called it "silly".
"They (Facebook) haven't misrepresented what we bought. So, you paid for impressions optimized for views. Both views and impressions were reported perfectly. It was only when they looked at viewed percentage then the problem cropped up," said Gautamm Mehra, Chief Data Officer at Dentsu Aegis Network.
However, the fact that an organization that accounts for a growing number of ad money when it comes to digital and, increasingly, video, made a mistake, no matter how small, will probably give critics enough fodder for a while and also raise questions about transparency once again.
To be fair, Facebook tied up with third-party agency Moat last year to independently verify video ads run on its platform. This came after a number of agencies and brands expressed their displeasure at having to take the company's word for things like viewability of ads.
"It is big news from a very credible organization. technically it is loss accrued by an advertiser and is necessary to estimate. Reading the fine print is necessary as only one of the metrics was overestimated. So it might not necessarily be the metric that advertiser RoI is measured on," said Shekhar Banerjee, COO of Madison Media Infinity.
"They did come out in public on their own but it is a rather irresponsible error in the first place. Considering that there is a lot at stake on the video front with YouTube and Facebook already has trouble with autoplay, this does dampen the spirit. However, the good news is that Facebook insights and remarketing capabilities are very robust today; in fact, superior to their competitors," further stated Mehra.
Ironically, the fact that Facebook only considered views of more than 3 seconds in its calculations actually would have provided marketers with a clearer idea of how much audiences are actually interacting with their content.
The general tendency in the industry seems to be to trust the big publishers like Facebook and Google when it comes to data they provide. Also, marketers may not want to spoil relations with publishers of that stature.
"If this was done by a start-up, brands and agencies would go on the offensive. But since it is Facebook, it will be interesting to see how soon the controversy will die. It may not make any difference to future spends on Facebook but it takes courage to apologize publically. On the other hand, maybe they know that they can get away with this," opined Madan Sanglikar, Co-founder & Managing Partner (South East Asia), Affle.