Friday, February 16, 2018

Facebook's Bullshit Boost Campaigns

I made this fairly nondescript post to the Facebook page for my app, "Eat Everywhere".



As always, Facebook offered to "boost" the posting for $10. Even though it wasn't carefully constructed, and didn't really work as a standalone sales pitch, I figured what the hell. I've wasted $10 on greater frivolity.

What happened was very interesting. FB claimed the posting was seen by 202 people. It was "liked" by 54: one real human being (who co-edited the app) and 53 ciphers. Many appear not to be English speakers, most couldn't pass a social media Turing test, and none seem like they'd have the remotest interest in the app....and certainly not this chatty vague posting. What's more, Facebook said that they'd confine viewing to USA residents. Uh-uh!



Consider our new fan ‎سیدعباس‎. Here's his account. Does he seem like a fully-fleshed out person to you, much less someone who'd remotely be interested in my app? Continuing down the "like" list, how about Tran Muon, who could not possibly be more sketchily etched, or more unlikely to "like" this posting? What about 陽菜?

Go through a few more, like Brian Omes and Donna Ramirez and Crystalon Cryer, and you'll sense a pattern. They all have friends, but those friends' accounts are equally stillborn, random, and weird. The pattern is consistent: six to ten photos, few or no actual postings, and a few dozen friends who appear to be in comically different movies. Not one is somoene I'd expect to like the app, much less a vague posting about that app.



Throughout this supposed "campaign", there were no new hits to the Eat Everywhere web site, nor downloads of our iOS or Android apps.

I'm assuming most people who buy these $10 boosts are quite happy with a few dozen "likes" - the Mardi Gras beads of social media ("Ditzcoin"?). People engaged in actual business wouldn't be, but, then again, they wouldn't dabble in these micropayment dangles. So Facebook recruits a mixture of fake people (i.e. bots) with fake accounts, real people with fake accounts, and real people with real accounts to push "like" buttons, and the "client" gets the handful of Mardi Gras beads they hoped for.

I didn't expect much for my $10. But would it have been so tough for FB to zero in on, say, food lovers, when the app's title is so easily parsable to their algorithms? If not, I guess I understand why they can't simply play straight and show 202 real people. 202 actual people will not yield any tangible result. 202 people would not be offering me these 53 Ditzcoins. I mean, one could offer a bona fide update from Jesus Christ himself to 202 random people without drawing more than a single "like" or two....if even that.

But I don't understand how the hell they get away with this. It's so incredibly flagrant!


See followup here

No comments:

Blog Archive