Facebook discloses more engagement measurement errors; some of the faulty stats were used by The New York Times and BuzzFeed in their stories about fake news
Facebook misreported engagement numbers for links and live videos and is refining the ad reach estimates it shows advertisers.
Context & Ripple Effects
This disclosure lands a month after Facebook admitted to miscalculating Organic Reach and Instant Articles time-spent and formed a Measurement Council under third-party verification pressure. The new wrinkle: the faulty link and live-video engagement numbers weren't just an advertiser problem — The New York Times and BuzzFeed cited them in their own reporting on fake news, meaning the errors contaminated editorial coverage of the very misinformation debate Facebook was under fire over.
First-order effects
- The New York Times and BuzzFeed now have to answer for stories built on stats Facebook itself says were wrong, while advertisers receive revised ad reach estimates mid-campaign.
Second-order effects
- Every subsequent error compounds the credibility tax on Facebook's self-reported numbers — by May 2017 the tally reaches ten mistakes since September 2016, forcing refunds to advertisers (the tenth disclosed error) and pushing buyers toward independently verified measurement.
Third-order effects
- The pattern ends in court: advertisers allege Facebook sat on a key video metric error for over a year (the 2018 lawsuit), pointing toward a structure where platform-reported metrics are treated as marketing claims rather than auditable data.
The trend: Platform self-reported metrics are steadily losing default-trust status, with each disclosure pushing advertisers and publishers toward third-party verification.