Facebook came under fire last week after their vice president of business and marketing partnerships, David Fischer, wrote a post apologizing for the company’s use of a miscalculated metric that incorrectly measured the time viewers on the platform spent watching videos.

“The metric should have reflected the total time spent watching a video divided by the total number of people who played the video,” said Fischer in his post, “but it didn’t.” 

Although this response came within the last week, the Wall Street Journal reports Facebook had quietly made changes to their “Advertiser Help Center” page addressing the metric about a month ago. The update acknowledged that over the last two years, Facebook’s “Average Duration of Video Viewed” measure had failed to calculate any video views that were less than 3 seconds. As a result, average watch times of videos were inflated by up to 80 percent.

This recent revelation isn’t the first time that Facebook has come under fire for miscalculated metrics. In 2015, Facebook announced that a glitch in their iOS apps inflated the actual amount of time users spent on the platform. While the glitch didn’t end Facebook’s relationship with advertisers, it did make them more cautious when relying on the platform’s in-house analytics.

Some users and advertisers have also previously questioned the metrics Facebook provided them. In August of last year, YouTuber Hank Green accused Facebook of deceptive statistics in a blog post entitled “Theft, Lies, and Facebook Video.” The post, which challenges a “claim” that Facebook is now streaming more than YouTube, highlights independently analyzed quantitative measures to show that Facebook’s three-second target inflates viewer rates.

Although it may seem to some like a “victimless crime,” Green says the implications for video creators who publish on Facebook are tremendous.  “It fundamentally devalues the #1 metric of online video,” Green’s post says, “Applying that word (view) to something far less valuable is going to be extremely disruptive to creators.”

3) I now regularly see our videos with 50MM+ view counts that are stolen by individuals on FB… sometimes by other media cos

— George Strompolos (@gstrompolos) June 4, 2015

While Fischer’s announcement isn’t likely to end Facebook’s popularity with digital advertisers or content creators, it will likely see some pressure from advertising firms to make its analytics methods more transparent. Although there weren’t any additional fees for promoters as a result of the inflated figures, the Wall Street Journal reports that too-good-to-be-true metrics likely swayed advertisers from choosing alternative platforms  for their media.

As a “walled garden” platform, Facebook is entirely responsible for how it both displays content on its site and measures the same content’s performance. For many, this seems to give the platform an inherent conflict of interest when it comes to reporting numbers to potential advertisers. When advertisers rely on metrics from platforms without third-party verification, they are also relying on the sites to report those numbers without manipulation for profit’s sake.

“Without third-party verification, it’s letting [platforms] mark their own homework,” said Keith Weed,  chief marketing officer of Unilever.

While Facebook has reportedly fixed the discrepancy, it has acknowledged the need to improve its analytics metrics and has recently announced new measurement partnerships. “Our promise is we will and need to do better,” Carolyn Everson, Facebook’s Vice President of Global Marketing Solutions, said recently on the issue.

But on the topic of third-party metric verification, Everson has a clear idea of where Facebook stands. “Make no mistake,” she said, “we don’t believe in simply grading our own homework.”

Lauren Shiplett contributed reporting.