Facebook discloses more ad-metric reporting glitches

LOS ANGELES — Facebook said it has discovered four additional user-engagement metrics that it has been misreporting to marketing clients.

However, Facebook says none of the errors affected the ad rates that marketers paid. By announcing the mis-stated measurements — and promising to fix them — the company is trying to show it’s taking good-faith efforts to be as up-front as possible with Madison Avenue. The social giant, which has more than 1.7 billion monthly active users, said it plans to work more closely with third-party measurement providers including Nielsen to verify its data and is forming a customer task force on measurement.

The update comes two months after Facebook revealed that it had overstated the average time spent viewing videos for two years. While that also didn’t affect ad rates, it gave marketers a skewed view of how well their ads performed.

“Our goal going forward is to communicate more regularly about our metrics, so that our partners can focus on doing what they do best — serving their customers — with the best insights possible,” Facebook said in a blog post Wednesday (Nov 16).

Here’s a summary of the metrics Facebook says were misreported:

PAGE INSIGHTS :

Facebook said a bug in Page Insights caused summary numbers showing seven- or 28-day organic page reach to be miscalculated as a simple sum of daily reach instead of de-duplicating repeat visitors over those periods. According to the company, the de-duplicated seven-day summary in the overview dashboard will be 33 per cent lower on average and the 28-day summary will be 55 per cent lower. The bug has been live since May; Facebook said it will fix it in the next few weeks. In addition, Facebook said it will change how it reports Page organic reach to match the way it tabulates paid reach: Reach counts will now be based on the stricter definition of viewable impressions (counting reach once a post enters a user’s screen, whereas it previously defined reach as a person refreshing his or her News Feed and the post being placed in their feed). With the stricter definition, Facebook estimated that reported reach will be 20 per cent lower on average.

VIDEO-MEASURING COMPLETIONS:

In cases when audio and video tracks don’t line up, audio in a video may continue to play for a bit longer, which caused Facebook to undercount the metric “video watches at 100 per cent” (previously called “video views to 100 per cent”). By updating how the company reads video length, publishers may see a roughly a 35 per cent increase in the count of “video watches at 100 per cent”, Facebook said.

INSTANT ARTICLES - TIME SPENT:

Average time spent per article had been over-reported by 7 per cent to 8 per cent on average since August 2015. Facebook said that was because it was erroneously calculating the average across a histogram of time spent instead of reflecting the total time spent reading an article divided by its total views. That’s now been fixed.

ANALYTICS FOR APPS — REFERRALS:

Facebook said it miscalculated a metric that evaluates all posts produced by people via an app or website. “We meant to count clicks that went directly to an app or website; however, we’ve also counted other clicks on those posts via the app or website, including clicks to view photos or video,” the company said. Out of the referrals it currently reports, on average about 30 per cent are actually clicks to consume content on Facebook; this glitch is in the process of being corrected.

Facebook also announced that because of low usage, it is phasing out “interest lists”, which have provided a way to organise and view content on the service. That will result in a drop in the total number of followers for profiles that created interest lists or were featured in lists. Most profiles will see a drop in followers of less than 5 per cent, the company said.

To boost its credibility with advertisers, Facebook is looking into working with additional third-party measurement providers to validate its metrics, in addition to Nielsen, comScore, Moat and Integral Ad Science (IAS). Facebook said it will provide viewability data from Moat, IAS comScore to verify the amount of time ads are viewed on-screen. It’s also partnering with Nielsen to include Facebook video viewership in Nielsen’s Digital Content Ratings (DCR), which promises to give media companies a comparable look across digital and TV metrics.

In addition, Facebook said it has created a new internal review process “to ensure our metrics are clear and up-to-date as our product offerings continue to evolve”, which will include communicating to clients via regular updates on its new Metrics FYI blog. VARIETY.COM/REUTERS