As detailed last week by Econsultancy’s Rebecca Sentance, Facebook’s faux pas with metrics it supplied to advertisers and publishers are causing new headaches for the social networking giant after a group of advertisers suing the company are claiming that the magnitude of Facebook’s errant algorithms, which inflated a number of metrics, was much higher than previously reported.

Facebook claimed that its inflated average viewership metrics by 60% to 80%. But according to advertisers who filed a lawsuit in 2016 and have since been able to review over 80,000 pages of internal Facebook records, the social media giant actually overestimated those metrics by a whopping 150% to 900%.

The impact could be more significant than expected at first glance. As Heidi Moore, a digital media consultant who used to work for Wall Street Journal and Guardian, sees it, Facebook’s mistake, combined with its practices, had a destructive effect on news publishers.

As Sentance explained, the situation is a lesson to brands:

“…marketers (and advertisers, and publishers) need to remember to think like regular consumers, and talk to regular consumers – especially in age demographics not their own – in order to keep grounded in what they want and what they’re doing, rather than letting themselves be blinded by the latest glamorous fad or impressive metrics.”

More data and metrics, more problems?

In addition to reminding companies to think like the people they serve, the situation also offers a case study in the perils of making big decisions based on now-abundant data and metrics that come from third parties.

Companies have more data and metrics than ever before. Some of the data is collected directly from their own websites and mobile apps, but much of it comes from third parties. This is especially true when it comes to social media. From Facebook and Instagram to YouTube and Twitter, companies typically have little to no ability to access the raw data associated with interactions their accounts and content generate on these companies’ platforms.

That data is used to generate metrics that businesses rely on to gauge the productivity of their efforts on popular platforms. Obviously, the inability to see raw data and audit the processes and algorithms that use that data to produce metrics, many of which are proprietary/non-standardized, leaves companies in a difficult position.

In some cases, companies do have the ability to test the integrity of the metrics they’re being provided. For example, a business that is running ad campaigns might have the ability to compare third-party and internal click metrics. This is of course how click fraud is typically detected.

But in many other cases, such as metrics for content views and engagement that take place solely on third-party platforms, companies have no choice but to trust that the metrics they’re being provided are accurate.

The Fundamentals of Marketing Measurement and Analytics

As the Facebook case demonstrates, this forced blind trust can be very problematic.

Complicating matters is the fact that platform providers can and do make changes to the way their metrics are calculated. For instance, YouTube just announced that it’s changing its attribution criteria for TrueView for action video ads. Going forward, an Engagement event will consist of an ad view of
10 or more seconds instead of 30 or more seconds and a Conversion event will consist of an action taken by a user within 3 days of viewing an ad instead of 30 days.

YouTube says that the changes were based on “large-scale experiments” and while they might indeed be sensible, they highlight the fact that companies are subject to the whims of the platforms they use.

The importance of auditing

The good news is that when it comes to the veracity of metrics, change is afoot. Facebook, as a result of its metrics mistakes, last year agreed to a Media Rating Council audit and has partnered with third party verification firms that companies can use to verify metrics such as video ad viewability.

But there are many metrics that are still not subject to third-party verification. This is problematic when one considers that various non-ad related metrics are being used to make sweeping decisions about strategy and investment.

This highlights the importance of establishing KPIs that can be measured directly and without total dependence on unaudited third-party data and metrics. Had more publishers done this, they likely would have been largely unaffected by Facebook’s faux pas.

Standing up for what matters

There is also evidence that larger companies can cut better deals for themselves when it comes to their investments.

Take Chipotle, for example. As AdAge reported, the US restaurant chain, through its agency, Mediahub, was able to negotiate a guarantee from Turner that its ads on Turner’s networks, including TBS and TNT, would deliver a sales lift. If they don’t, Turner will issue make-goods.

According to AdAge, other networks are pursuing similar models and another network, A&E, has inked five deals that promise specific outcomes.

While television and cable networks are in a more difficult spot than the Facebooks of the world, and thus have less negotiating leverage, these deals should remind companies of the virtues of standing up for themselves and ensuring that they don’t lose sight of what they’re really trying to accomplish instead of getting distracted by data and metrics that might or might not be accurate.

Learn more

Econsultancy offers measurement and analytics training courses