Thursday, April 27, 2017

Brand Safety Issues #2 - Facebook's Issues

As suggested in the last post, Facebook has had its own brand safety issues in recent months. Some of them are different from Google’s; one is similar.

Problem #1. The recent set of issues began in September 2016 when the Wall Street Journal publicized the fact, already revealed by Facebook, that it had been overstating one of its key video ad metrics for two years. The metric was average viewing time, highly important to video advertisers. An audit of all 220 Facebook metrics revealed four other problems with specific metrics.  In December the platform revealed that three of its audience measures had been calculated incorrectly.

The Resolution. Facebook began releasing fixes including individual metrics later in the fall.  It also announced initiatives that included more third-party metrics, a Measurement Council with members from ad agencies of metrics companies, and, in general, more transparency about how its metrics are produced.  In general, metrics are likely to provide additional problems from time to time and produce new metrics or new definitions of old metrics. This reinforces the warning at several points in the book, especially in the SMM and metrics chapters, that marketers must read and understand the metrics definitions of each platform they use—on an ongoing basis, it seems.

Problem #2. Fake News became a major issue during the 2016 election campaign, although all journalists agree that it is a persistent issue. I’ve seen references back as far as the ancient Greeks! Whenever the first fake news stories were circulated, there is no doubt that social media has given them a huge megaphone. What can—and should—social media platforms do to combat false stories masquerading as news? (See Related Links in previous post for Google actions.)

The Resolution. During the election there were claims that Facebook’s Trending news feature was left-leaning and Facebook replaced the human curators with an algorithm. That sounds like a different issue,  but part of the problem was that biased and fake stories were being shared and therefore showing up in the Trending stream.

The scope of the problem became obvious before the French elections in Spring 2017 when Facebook shut down 30,000 fake accounts that it claimed were circulating deceptive content. Before the surprise British elections to be held in June 2017 Facebook was being pressured to do more to stop fake news. It told the Guardian that it had already upgraded software as well as using human fact-checkers.

   https://thenextweb.com/facebook/2017/03/06/facebook-can-now-flag-fake-news-stories-disputed/#.tnw_6yGFqabv


Among the general steps it has taken, Facebook has partnered with third-party fact checkers and recently began flagging disputed stories. It also added a tool to make user reporting of questionable content easier.   It has also taken steps to keep sites with false or offensive content from hosting advertising.

Problem #3.

Facebook has the same problem of ads appearing on inappropriate sites as does Google/YouTube although Google’s problems dwarf reporting of the same issue on Facebook   If CNBC is correct and this problem centers on video, then Facebook’s problems in this area are likely to increase.

The Resolution.

Facebook’s response is similar to that of Google. It is trying to keep ads off fake news sites. Facebook has partnered with third-party face checkers and begun its own project to fight fake news. It has also joined the News Integrity Initiative which is to be run out of NYU’s journalism school.

None of these problems is prone to easy or quick solutions. The trust in two major social platforms has been eroded. It will be hard to reclaim it even if nothing else happens. Given the state of the world at the moment, nothing else happening seems to be an unlike scenario.

Before I leave the fascinating subject of digital advertising for now, I want to acknowledge the other huge presence that is just emerging. So look for a post on Amazon advertising soon.

Related Updates

Disney introduces safe brand network; partners with Facebook
How to identify fake news with Facebook criteria
Superposters help identify fake news
Facebook and Twitter have tools that show the ads for a particular account
Facebook email dump
Timeline of 26 Facebook data blunders in 2018 
Facebook sets aside $3b for potential fines

No comments:

Post a Comment