Facebook is facing renewed scrutiny over the way it reported its video metrics, after an updated lawsuit was filed in California this week alleging the social media giant’s errors were much worse than originally reported.
The original filing was made in October 2016 by a group of small advertisers after Facebook admitted it had miscalculated video viewership on its platform by 60 to 80 per cent for two years.
That complaint stemmed from a Wall Street Journal report in September 2016, that revealed Facebook had been overstating the average time its users spent watching paid video advertisements.
In response Facebook admitted the mistake but maintained it only discovered the error one month before and had fixed it the errors.
The amended class action now alleges that Facebook overestimated ad viewing time by 150 to 900 per cent, and that it knew about the issue for a year.
The court documents allege Facebook was deliberately slow to fix the problem and “developed and deployed a ‘no PR’ strategy designed to ‘obfuscate the fact that we screwed up the math’.”
The plaintiffs are seeking punitive damages and argue Facebook’s “dissemination of inaccurate and inflated video advertising metrics” constitute fraud.
Facebook says the lawsuit is without merit and is seeking to have the claims dismissed. Telling WSJ, “Suggestions that we in any way tried to hide this issue from our partners are false. We told our customers about the error when we discovered it—and updated our help center to explain the issue.”
The plaintiffs argue that the inaccurate figure induced advertisers to spend more on Facebook video advertising than they would have otherwise.
As a response to industry concerns over the way it reports its metrics, Facebook has been working with third party measurement companies and the Media Rating Council (MRC) to audit its data.
In June Facebook had 24 partners in its measurement system, such as Oracle Data Cloud, Nielsen, Kantar Millward Brown, as well as three partners measuring viewability, Moat, Integral Ad Science and ComScore.