Facebook: more metrics mistakes

Following 2016’s Facebook video metrics scandal, new problems with the social giant’s data have surfaced: Facebook has revealed that data reported for Instant Articles is wrong. Read in Italian.

In an official post  published in its newsroom, Facebook sates:  

We’ve uncovered an issue for a small group of Instant Articles publishers that impacts reporting in comScore […] This caused an underreporting of iPhone traffic from Facebook in comScore products between Sep 20 to Nov 30, 2016. iPad and Android traffic were not affected.

In brief, the social network claims that it incorrectly reported the numbers of iPhone users reading its content – to the detriment of publishers and marketers.

Instant Articles are multimedia posts used by many publishers because they let readers access content faster, without having to leave Facebook to load them on an external website. In return, publishers earn money through advertising inside the Instant Articles. Thus, calculating the exact number of times the content is accessed, and consequently the traffic generated, is crucial.

The Menlo Park giant has always claimed that mistakes concerning metrics don’t affect invoicing. However, the problem is that advertisers need metrics to understand users’ actions and behavior on the platform, so they can decide when and how much to invest. Consequently, they must not be neglected.

This time, since the reported figures were lower than the actuals, many publishers were negatively impacted. Among those reporting the biggest differences were The Washington Post, BuzzFeed, Mic, Entrepreneur, Foreign Policy, Inverse, PopSugar and Variety.

So far Facebook has managed all metrics “in house”. It’s about time that it opened up to third parties to try and achieve more objective assessments. In this respect, in addition to comScore, the social media giant is also partnering with Integral Ad Science and Moat.

By the way, recently, even the Media Rating Council (the United States Congress audience measurement regulator, set up in 1960 to check whether audience measurements are valid and respect certain criteria) asked for more clarity.

Ensuring that external bodies confirm figures provided by a platform boasting 1.6 billion users are reliable, not random, is a hot topic. In Italy, we are also facing this issue.

Recently Audiweb signed several agreements with different publishers, including Century Fox, Caltagirone Editore, Conde Nast Digital, Gruppo De Agostini, Gruppo Espresso, Il Sole 24 ore, Mediaset, Mondadori and RCS MediaGroup. Alongside Nielsen, Audiweb has been monitoring data on content consumption inside the mobile Facebook app,  including browsing and/or Instant Articles, so that it can provide verified, useful data to media and advertisers.

What is your opinion about metrics published by social networks? Do you think they’re reliable?  I would like to get your comments. Tweet to @agostinellialdo.

If you liked this post, you should read: Facebook video metrics: what to believe?