Facebook’s video-ad scandal shows disdain for advertisers, users.
Updated: Dec 14, 2019
This piece first appeared as an OpEd on DowJones MarketWatch
The takeaway from Facebook’s bombshell that it had overstated a key measure of video ad viewership on its pages by about 80%: It’s not that the social-media giant “miscalculated” its numbers. It’s a fresh view into Facebook’s arrogance, both toward its advertisers and its nearly two billion users.
Nothing’s new under the sun, of course — pride has goeth-ed before a fall for all of human history. And obviously does in the digital sphere, too: Facebook FB, +0.36% and its principal digital advertising rival, Google GOOG, +0.31% have been behaving as though customs and principles that have governed the advertising business for its entire history simply don’t apply to them.
Businesses based on selling time or space to advertisers have almost always allowed third parties to measure their effectiveness at obtaining eyeballs. The print-publishing industry had the advertiser-created Audit Bureau of Circulations looking over its shoulder, while the broadcast and cable industries had Nielsen and Arbitron ratings that independently determined viewership of TV and radio shows. This “third-party verification” setup is just about the only way to get reliable data without self-interested bias or spin.
You’d think this matter could be easily solved with a big dose of transparency.
That kind of oversight on behalf of advertisers is especially critical for platforms like Facebook and Google’s YouTube, which are so-called “walled garden” sites. That means you usually view content created by other media companies and individuals without leaving Facebook or YouTube — the content plays inside their “fence.” Which, in turn, means Facebook and YouTube are in complete control of all the data regarding clicks, views, demographics and other advertiser-coveted data.
But the two giants have only recently made begrudging steps toward this kind of third-party verification. Facebook started allowing analytics firm Moat to begin a limited beta test on its site this year, while YouTube announced it would allow certain “viewability vendors” to begin to monitor its video ads. But they did so only under heavy pressure from major paying customers, who were tired of bankrolling multi-billion dollar digital ad budgets without knowing exactly what was and wasn’t being seen by consumers.
And Facebook and Google also didn’t give the new third-party monitors full access to monitor and measure whatever they wanted to. They can only verify, for now, limited ad viewability data.
Watching this unfold feels a great deal like an Iranian nuclear program negotiation: Both companies still have their hands wrapped tightly around the neck of the outside monitors, the access to information is far from complete, and in the end you can’t help but be left with the feeling that both companies are hiding something.
Why is this one Facebook “mistake” so telling? Probably because the metric is so key to digital advertisers. Video advertising is largely not used to drive direct responses; it’s used as air cover for building brand image and awareness. If you’re an advertiser trying to measure return on your investment, the length of time consumers are spending watching your digital video ads is about the best proxy you can get for an effectiveness index.
When advertisers can’t calculate or gather that data from independent sources because the social-media site they’re buying ads from won’t be transparent, and when that same site then turns around and says, “whoops — we have been overstating your key index by 80% for the past two years,” that’s a major problem in the digital-advertising world.
What’s worse is that Facebook’s claimed engagement numbers were orders of magnitude higher than any other platform. Facebook claimed superior targeting as the reason, asserting they had simply created a vastly more effective marketing machine. But in the light of day, that comparison now looks more like a 2007 Bernie Madoff investor presentation than it does evidence of a better mousetrap.
Like those frustrating nuclear negotiations, you’d think this matter could be easily solved with a big dose of transparency. And like those nuclear negotiations, you’d be forgiven for concluding it’ll never happen. Facebook’s response has been to rename and recalculate the metric on which they were caught cheating. Sleight of hand still rules.
No one knows exactly how the “mistake” was discovered, or why Facebook was excluding viewers who clicked away from the video at the first seconds of the ad (artificially boosting the average viewing time). The beta implementation of the third-party platforms clearly could have led to it – and if that’s the case, what else will be uncovered as the waters of obfuscation recede?
If they recede, that is. Both companies have well-deserved reputations for doing what’s best for them, and advertisers and users alike can like it or lump it. What are the take-away lessons from all this inside-baseball controversy?
First, like in all the rest of the world, numbers that look too good to be true usually are too good to be true. The biggest media platform in the world was claiming its video-engagement performance was vastly better than every other media site’s, flashing performance numbers that seemed ridiculous to informed observers. So either A. Facebook believed its own hype, or B. They knew their numbers were juiced and kept on using them for as long as they could.
Second, digital advertising continues to be the Okefenokee Swamp of modern marketing, with virtually no visibility and plenty of danger lurking in the muck. The uncorrected average Facebook video view duration was reported by one source last year to be about 18 seconds; an 80% “error” would put that average view duration at just 10 seconds. Meaning there’s a huge amount of wasted ad spend even in so-called “performance marketing.”
Meanwhile, adblocking continues its unabated growth, with media sites desperately trying to figure out what policies or work-arounds or adblocker blockers to deploy.
What’s needed to ameliorate this mess is full transparency to the advertiser about how their media behaves. Programmatic ad buying, and in fact all digital advertising, has a lot of warts on it. It’s not like marketers are unaware of the fact, and yet many media platforms want to pretend it’s not true. The first step in wart removal, of course, is admitting you’ve got a wart. If the industry would take that simple, yet profound, step of transparency, its path toward a more stable, bad-surprise-free future would be greatly enhanced.
Real data and metrics can help guide marketers to make better ad placements. Better placements, in turn, help align ad content with consumers who actually want and need to see it, which means they’re more valuable to the advertisers and therefore capable of generating more revenue for the media site.
Which, in turn, could once again begin to build a healthy content ecosystem.
That’s a much healthier dynamic than having 90% of us angrily poking away to close an invasive mobile ad that’s trying to seize control of our smartphone.
This is a seminal moment in a business that’s really only existed for seven years. If advertisers band together to demand a more open, less high-handed set of behaviors from Facebook and Google — who between them have created almost a trillion dollars of market capitalization — the expectations for performance and transparency can be reset. If a stand isn’t taken here, the beatings are likely to get worse.