Just weeks after Facebook revealed it had miscalculated several video metrics, it announced it had found more measurement errors. In a blog post, Facebook reported two measurement errors.
It found an inconsistency in how it counts engagement such as likes and shares with off-site links in its mobile app. Specifically, it identified a discrepancy between the counts for the Like and Share Buttons via its Graph API and the counts when users entered a URL into the search bar in the Facebook mobile app. It is working on resolving the issue.
Facebook admitted it had misallocated the reporting of multiple Reactions from unique users during live video streaming. It attributed extra reactions per user during live broadcasts to the “Reactions from Shares of Post” section, instead of counting them in the “Reactions on Post” section. The fix for the issue will begin in mid-December. “Reactions on Post” will increase by 500% on average and will decrease on “Reactions from Shares of Post” by 25% on average.
In addition, Facebook said it is changing how it calculates the number of people who potentially see ads. The new methodology relies less on projecting data using smaller samples of Facebook audience. This change should decrease or increase Facebook’s ad-reach estimates by less than 10%, it said.
Previous Overestimates in Organic Reach
Previously, Facebook revealed it overstated organic reach in its Page Insights since May, over-reported average time spent on Instant Articles since August 2015, had overstated the referrals metric in its Analytics for Apps dashboard, and had substantially overestimated the average time users spent watching videos.
PR pros, marketers and advertisers may be confused and perturbed at the numerous errors and continuing adjustments. At least some appreciate Facebook’s transparency and recognize the difficulties of reporting many metrics for new products. Admitting mistakes, correcting them and increasing transparency will create more confidence in Facebook’s measurement over the long run. Yet marketers and advertisers may hold it to a higher standard because of its dominance in social media dominance.
No Margins of Error Please
“They are trying to get it right,” Ian Schafer, founder and chairman at ad agency Deep Focus, told the Wall Street Journal. “Yet they are a data business, with a huge responsibility. And advertisers don’t like margins of error much.”
Brands may increasingly pressure the company to allow more direct access to metrics and third-party reviews of data. In an earlier blog post, Facebook said it is exploring additional third-party reviews to validate its reporting.
“Though none of these problems are mission-critical, they are further examples of Facebook having issues with its systems for providing transparency to those who use the network to make business decisions – whether that’s where and how to spend their ad budgets, or how to best grow their social media audience by using Facebook’s products,” says TechCrunch writer Sarah Perez.
The Facebook issues illustrate the enormous difficulty of getting communications analytics exactly right and the need for both suppliers and clients to continually check and improve data collection, metrics and methodologies for aggregating analytics. When suppliers admit and correct data and analytics errors, that’s a good thing for clients. Valid data help assure sound insights.
Bottom Line: News that Facebook miscalculated several metrics probably raised concerns among marketers. However, the company’s forthcoming statements and efforts to resolve the issues likely calmed fears. Its plans to increase transparency may help increase confidence in its measurement practices over the long term.
William J. Comcowich founded and served as CEO of CyberAlert LLC, the predecessor of Glean.info. He is currently serving as Interim CEO and member of the Board of Directors. Glean.info provides customized media monitoring, measurement and analytics solutions across all types of traditional and social media.
Trackbacks/Pingbacks