Did Facebook really rip off its video advertisers?

No. It is a complete non-story - hyped by limited research and hyperbolic headlines, suggesting Facebook did something fraudulent. Journalists have been quick to jump on the bandwagon, all feeding off the initial story without digging into what it’s about.

What actually happened?

Here’s some context to help break this down into what it really is.

Facebook only charges marketeers when users watch their video ad for 3 seconds or more. They use a metric called ‘Average Duration of Video Viewed’. Any right minded person would naturally assume this only included those ads classed as a ‘view’.

Unfortunately Facebook’s documentation was wrong:

  • Previously the Average Duration of Video Viewed was equal to: total time spent watching a video, divided by the total number of people who have played the video.

  • Now the Average Duration of Video Viewed is equal to: the total time spent watching your video divided by the number of views of your video, for three or more seconds.

Realistically it would have been a rather poor marketer who based any further buying decisions on this metric alone instead of looking at the raw data. Facebook have not changed any code to correct this, only their documentation - so caution and careful consideration must still be applied when considering your purchases.

Is Facebook advertising completely flawed?

Let’s not get carried away too.

I’ve said it before and I reiterate, Facebook’s news feed ads, fed by its data & attribution products is the greatest ad format ever invented. Marketers can easily see if they are getting ROI, and attribute actions users take directly back to an ad they’ve seen. If they’re not seeing results, they wouldn’t buy more.

The real story

The real story isn’t a metric miscommunication or blunder - it’s a much bigger issue. The debate here is that of walled gardens, does Facebook provide enough transparency for independent verification. It’s recently partnered in a limited way with Nielsen, Integral Ad Science and comScore to provide some transparency, but there are >20 others on agencies wish list too.

For Facebook it’s a question of user data, that’s their USP. So controlling it for themselves and making their users feel protected so they continue to spend time in their Apps is key. Opening their door to too many vendors, and knocking down their walled garden unquestionably puts that at risk.

If you want to read Facebook’s explanation of what happened you can find their attempt to clarify it all here, in the Facebook Business help centre.