There I was, just minding my own business on Twitter, when I saw this tweet that finally illuminated for me the nexus of at least one part of the presumed “collusion” between the Trump campaign and the Russian government-backed troll factory.
Although we already had a vague idea of how Facebook and other social media became the space in which the collusion lived, ironically it wasn’t until Wired published its piece, “How Trump Conquered Facebook—Without Russian Ads” that I finally came to grasp the Russian role in the online coup, and exactly why the Trump campaign was always such an unrelenting stream of hot, blazing trash.
The basic premise of the Wired piece is that the still-mostly-unregulated Facebook advertising market works in ways that the traditional media advertising market cannot. This is true both in terms of the hyper-granular analytics that only a trackable, web-based market can provide, and in the fact that the lack of regulation frees Facebook and other social media to offer different effective ad rates to different campaigns—a practice prohibited by law in the traditional media.
Another point of departure between social and traditional media is the emphasis social media places on engagement. It’s the mother’s milk of social media networks. Engagement signals attention, and eyeballs. It’s the metric through which social media networks can convince advertisers of the value of their ad space. Traditional media sell you with circulation numbers. “I don’t know how many people will actually notice you ad, much less how many will do something with it. But I can tell you how many can potentially see it, and the answer is: our circulation number.”
Online media, on the other hand, can measure how often an ad comes up on screen, and they can be reasonably certain that that screen is where the reader’s eyes are. And with interactive, clickable content, they can measure exactly how “engaging” that content is. With “share”-able or “like”-able content, online social media can measure how frequently “engaging” content causes users to “organically” push that content to other users. You’re doing social media networks a tremendous favor by showing other users what you “like” and want to “share” with them. Hell, you’re doing their jobs for them. And you deserve a reward for it. A reward that comes in the form of lower effective costs for advertising that does them this favor.
One problem, though, with valuing “engagement” above all else—and with using a computer algorithm to measure it in real time, or even estimate it ahead of time—is that there’s no room for value judgments about the content, or what makes it “engaging.” And that means that content that’s “engaging” because it’s super-informative, accurate and penetrating is regarded no differently than content that’s “engaging” because it’s outrageous, infuriating and total bullshit.
As you might suspect, that’s a problem for, well, everything on which we have previously believed our society should be based.
On social media, then, where more and more of us are getting our “news” every day, the more "provocative" the content, the more “engaging” the Facebook algorithm considered it to be. And the way that potential engagement was rewarded in their advertising market was with lower effective pricing. In other words, the more chock full of outrage-inducing bullshit the content the Trump campaign looked to pollute Facebook with, the less Trump's campaign had to pay for dumping it there. That’s outrageous and infuriating all by itself!
Meanwhile, we’ve also learned that being "provocative" was the chief aim of the Russian trolls rolling out their tailored streams of propaganda, and the targeting of their provocations was vastly improved by the information and opposition research they were able to hack out of the emails of high-level campaign operatives and national party computer networks. Those hacking and phishing expeditions eventually yielded top-quality provocation material, high in “engaging” content, which in turn drove the prices for the ads and other content exploiting that material way down. The more poisonous the click-bait, the cheaper it was to run it, and the more Facebook space it took up.
It should be immediately obvious to you that there’s something very wrong here. First, that the two campaigns were charged different rates for their advertising. That’s wrong on its face, at least in the world of traditional media. But it’s far worse when you realize how heavily Facebook rewarded the most poisonous content.
Apparently, it’s a matter of some contention just how substantial the rewards for provocative content actually were, even though there’s almost certainly a wealth of hard data on it. If you buy the estimate offered up by the Trump 2016 campaign’s digital director, Brad Parscale, the Facebook algorithm rewarded their “provocative” content with a price 100-200x less than Clinton’s boring “normal” campaign ads.
Parscale’s firm, Giles-Parscale, took in a reported $94 million from the Trump campaign for digital advertising services, a substantial portion of which was surely profit to the firm, leaving what we might conservatively estimate as several tens of millions of dollars for actual ad placements. At a 100-200x price advantage, that could look and feel online like several billions of dollars worth of digital advertising! In other words, Facebook would have built a machine that allowed the most poisonous, insidious bullshit ever to have billions of dollars worth of impact.
So, the more clicks, likes and shares this poisonous, insidious bullshit got, the lower the cost of the next ad buy would be. But it’s also been reported that the Russians themselves spent “only” about $100,000 on Facebook ads (though we don’t know what sort of force multiplier, if any, their poisonous, insidious bullshit enjoyed). Thus Wired’s headline, “How Trump Conquered Facebook—Without Russian Ads.” The piece still holds the door wide open, of course, for Trump campaign-paid ads conquering Facebook. But it doesn’t ask any questions about how and why the decisions to populate the content of those ads with very specific types of poisonous, insidious bullshit were made. We have to do that ourselves.
So, let’s do that. If, as an expert in the mechanics of digital advertising, you’re already armed with an understanding of what, in general, you can do to drive your ad costs way, way down (and by comparison, drive your opponent’s costs way, way up), you’re definitely ahead of the game. But understanding the mechanics alone doesn’t actually help you produce content that does the trick. That requires an understanding of the peculiar political sensitivities capable of driving Americans up the wall. And the stuff that’s best at driving us (or anyone) up the wall is the stuff that has, at its heart, some kernel of truth (or even just “truthiness”), but blown as far out of proportion and sensationalized to the maximum extent possible.
And nothing gives the impression of truth (or “truthiness”) quite like claiming that your key “facts” have been hacked straight out of the enemy’s own emails.
True, “truthy,” or not, targeted, poisonous, insidious bullshit can race literally around the world—at steeply discounted prices, we’ve now learned—if there is, let’s say, a government-sponsored troll/bot army providing an instant, coordinated boost capable of effortlessly launching it into “trending” status, or exploding into virality.
Well, gee whiz, it seems that happened, too!
So if you, like me, wondered during the campaign, “How could this horrible shit be working?” now you have your answer. The more insane the content, the more “engaging” it was, and therefore the lower the price to run it, and the wider its reach.
Being batshit insane, then, was literally the business model.
Of course, if you needed to get even more seriously insane with your material—so insane that you actually worried about the reputational damage from running it—you could always hide your tracks somewhat, by using Facebook’s now-infamous “dark posts.” But even that only helped drive the “engagement” factor, which helped drive prices down.
Facebook may be able to walk away from this in the end, I guess, by resorting to the aw, shucks dodge, “Hey man, we were originally just supposed to be a way to look at hot chicks and rate them online. We didn’t know Russian spies were going to do this.”
Well, maybe. Probably, even. But they did build this almost-perfectly awful algorithm. And they did spend big, big bucks lobbying and fighting tooth and nail against being regulated like a traditional media company. They went out of their way to protect a regulatory scheme that allowed them to charge different candidates in the same race vastly different prices for advertising.
And they did design that algorithm to reward provocation and engagement. That might have made sense at some point. But I doubt they’ll be able to convince those skeptical of their claims that their platform was but a small piece of the election puzzle that their data analysts didn’t know exactly what was happening in real time.
So, now you know why the guy without any other political experience is set to manage the re-election campaign of an incumbent president—a job that normally falls to a much more sophisticated political operative. One who actually has a keen understanding of the intersection of politics and policy. But if Facebook performance is literally all that matters, well, that’s how you’d end up making your digital advertising mechanic your campaign manager. You won’t have to understand the politics that drives the news. With the amount of money they’ll have to spend on the most provocative ads ever, there will be no "news," other than what they create.
This won’t be a campaign that needs management, coherence, persuasiveness, or diplomacy. In fact, the more of that “boring” stuff there is, the worse their campaign will do on Facebook. A Trump campaign manager only needs to know how to drive the algorithm to ever-lower pricing. And the answer is: shitposting.
For our journalist friends simply observing the Trump “presidency,” this means there will be no “pivot.” Ever.
The Trump campaign’s business model depends on creating “engagement.” And for them, that means more and ever-increasing outrage.
By the way, before I sign off: this story also serves as a preview of the likely effects of the end of net neutrality, come to life. From now on, even without an end to net neutrality, social media networks will be able to feed you far greater volume at far lower prices for garbage from a favored client, while offering only clunky, high-priced delivery for information from the client who doesn’t “pay” them by providing “engaging” content.