Facebook Mischarges Advertisers For The Tenth Time Since September

Facebook has mischarged their advertisers once again, making this the tenth time since September that the company has incorrectly calculated advertising metrics.

Marketing Land reports that Facebook charged advertisers incorrectly for video carousel ads that appeared on Facebook’s mobile site when clicked on by smartphone users. Facebook announced this error in a blog post published on Tuesday. Facebook is reportedly refunding all advertisers affected by the error.

According to Facebook, of the total ad impressions served by the company during the year-long period that this video error was operating, only 0.04 percent of impressions were affected by the glitch. The median amount of money an individual advertiser is being refunded as a result of the glitch is reported to be $10. This is quite a small glitch in comparison to issues that Facebook has faced in the past, but there are a few reasons for this.

The glitch only affected video carousel ads on Facebook’s mobile site. This audience is much smaller than Facebook’s desktop website, the size of Facebook’s mobile website audience became apparent last year when Facebook decided to fight ad blocking. The company opted to ignore mobile web ad blocking entirely, despite the ability to do so thanks to Apple’s iOS enabling developers, who developed software for iPhone and iPad owners to block ads on mobile sites such as Facebook’s.

The error also only occurred if a user clicked on the ad. Advertisers were only billed if they had agreed to pay per number of times Facebook users clicked on their advertisements. If the link was not clicked, the advertiser would not be charged. Advertisers that purchased video carousel ads based on the number of impressions or video views were not affected by the glitch regardless of whether the video was viewed on Facebook’s desktop or mobile website.

It seems that the source of the error was with Facebook’s understanding of a video click. Facebook charged advertisers for the number of times someone clicked a carousel video into fullscreen rather than if they clicked on the video to follow an advertisement. This is not unlike an error Facebook faced in November when the company revealed that the Facebook Analytics app had had incorrectly measured referral traffic sent Facebook to a developer’s mobile app or website. Similarly, if a user clicked a link to open a photo or video, Facebook counted these as link clicks even if the user never left the Facebook website.