Sheryl Sandberg was in New Delhi last week, doing whatever COOs do on those trips (a lot of meetings with government officials and a keynote at some conference, from what I read), and a reporter finally got her to say something about the Facebook mood study. Her answer, more or less: it was "poorly communicated," and Facebook needs to communicate better about the research it does. That's it. That's the statement everyone's been waiting on for a week and a half.
I want to be fair here, because I'm generally not one of the people who thinks every Facebook decision is a plot against the user. But come on. "Poorly communicated" is a phrase you use when you sent an email at 4:58pm on a Friday about a schedule change. It is not the phrase you use when your data scientists ran an experiment on 689,003 people's emotions in January of 2012, without telling any of them, to see if showing people sadder posts made them post sadder things themselves. The problem was never the writeup. The paper in PNAS is actually pretty clearly written, if you're the type who reads academic papers for fun, which I sort of am, God help me. The problem is the experiment happened at all and that a checkbox buried in the Data Use Policy is apparently what passes for informed consent around there.
PNAS itself seems to agree something was off, by the way. The journal's editor put out what they're calling an "editorial expression of concern," basically a formal note attached to the paper saying the ethics review was murky and that Facebook's internal process, not any university IRB, was what actually cleared the study. That's a pretty unusual thing for a journal to publish alongside a paper it already accepted. I don't know how often PNAS does this (not often, is my guess) but it tells you the backlash wasn't just Twitter being Twitter.
What gets me is the "we'll communicate better next time" framing, like the actual fix here is a better press release. Communicating better about running an unannounced psychological experiment on people using your product to talk to their friends and family doesn't make the experiment okay, it just means you tell people afterward with nicer sentences. I went and reread Facebook's Data Use Policy that night out of morbid curiosity (I hadn't looked at it since I signed up back in college), and the relevant line about "research" is one clause, tucked between stuff about advertising partners and data retention. Nobody read that and thought "ah yes, I am consenting to be part of a psych study."
Small tangent, but this is also just bad optics timing-wise. I wrote about Google's right-to-be-forgotten form back in June, the one that already had something like 12,000 requests in its first days, and the whole reason that story exists is because a lot of people in Europe are furious about companies making decisions with their data without asking. Facebook picking this exact stretch of the year to get caught running secret mood experiments is either terrible luck or a company that genuinely doesn't think about how these things land outside its own building. I'd bet on the second one, honestly.
None of this means I'm deleting my account, I still use Facebook to keep up with people I otherwise would've lost track of years ago, my cousin's baby photos aren't going anywhere else. But I did turn off a couple of the "improve your experience" data-sharing toggles buried in my settings that night, for whatever that's worth, which is probably not much. It felt better than doing nothing.
The part I keep coming back to is how normal this whole exchange sounded on stage. Sandberg said her line, the room moved on, and by the next day the story had already started sliding off the front pages, overtaken by whatever World Cup result happened that afternoon. Companies have gotten very good at waiting out a bad week. I don't think that's a conspiracy either, it's just math: enough news happens every single day that a non-apology only has to survive until the next thing.