Amazing article about Facebook's bad data that led to the mass "pivot to video." People were willing to believe FB's data because it was FB, even though there was contrary data, and the data itself flew in the face of all of their own experiences. niemanlab.org/2018/10/did-fa…
The faith placed in Facebook's data is kind of stunning. It makes me feel less sympathetic to the publishers who may have been defrauded. There were plenty of reasons to believe FB was full of crap and people were saying so at the time. Still FB really is culpable here.
When the "pivot to video" wave hit, so many of us were saying, "Huh? That doesn't make sense," because it didn't make sense, the idea that video was the way for information to be communicated more quickly (which was the rationale) was nonsensical. Should've led to ?ing of data.
I used to work in survey research and the most important skill was knowing when data might be lying to you. Identifying data that didn't make sense meant you dug deeper to see if it was real or illusory. About half the time, that thing that didn't make sense really was error.
My survey research experience was almost 20 years ago, before people fell into the thrall of algorithms, but if anything we should be more skeptical of data now, not less. There's more of it, there's more incentive to misuse it, and so much of it is not transparent and open.
We'll see how purposefully deceptive FB was being when it comes to this video data, but if anyone trusts a single thing coming out that company, I got a proverbial bridge for them.
