Not That Tall... Facebook's Bloated Metrics

Remember when we were tonguing praise for Facebook Video a while back?

Welp, there’s a little hitch.

The Wall Street Journal recently released the headline “Facebook overestimated video metrics by 60-80% for 2 years”, which isn’t exactly news to third-party analysts: we’ve known for a while that there is a skew in Facebook’s metrics because they record a “view” as a video playing for 3 seconds or more. To have Facebook personally point this out and apologise, is telling of how seriously skewed these projected metrics really are.

This is something we've highlighted with our clients in the past, in some cases creating unbelievable confusion and additional work to more accurately convey the true value of engaged views on Facebook relative to YouTube's more sincere approach: 30 seconds.  

So the greasy result is this: Facebook’s idea of a “view” conveniently made the quality of Facebook views seem better than their video-centric competitor, despite not being a fair comparison.

The result: we’re left with two years’ worth of Facebook video metrics that might be 60-80% over-estimated. More importantly for businesses: this is money that could have been mis-spent.

Our take on this oil-spill?

Well: public view counts are important. The vanity metric does hold value against crowd psychology - “hidden gems” don’t tend to dictate quality on the internet. It’s a number - THE number - you can brandish to show who’s (ostensibly) been paying attention to your content. Which is why it needs to be accurate.

The analytics of online video advertising are progressively shedding light on some of the key considerations about how video is consumed. ATL video may very well have some similar, problematic nuances which deserve scrutiny, but we just don't have the engagement data, or tools, to see the true picture… so to speak.

For such a powerful metric tool as views, it’s formidably painful research and money-time wasted thinking Facebook view counts are an indication of user preference and enthusiasm by obfuscating the metrics.

So - why not look at the metrics behind the 30 sec view? This is undeniably a better measure of engagement. Consider that Facebook videos auto-play - how long do you glance over them before you decide to scroll down and move on? In less time it takes you to decide to commit to a Youtube video, we bet. The other big concern for those interested in authentic engagement is the way Facebook currently handles sound. Are we serving muted ads?


As The Wall Street Journal outlines - there is a bigger than ever need for transparency and third-party tagging and metrics to be able to verify Facebook data instead of a skewed source serving users bloated figures.