Full Frontal Facebook

In time, the savage bull doth bear the yoke.

—William Shakespeare, Much Ado about Nothing

OCTOBER 9, 2012

For Facebook, circa late 2012 was a moment worth savoring. Caught between startup infancy and big-company adulthood, Facebook in the year of the IPO gleamed in a brief adolescence, never to be repeated.

What was it like?

<beep!>

Microsoft Calendar, that cruel taskmaster, reminding me of a meeting.

Since launching FBX in June with a limited set of partners, we had begun admitting as many as possible. Every additional partner was an additional set of advertisers, which was more money and budget and internal leverage for me to use in FBX’s favor.

Every meeting with a potential FBX partner essentially boiled down to me taking a meeting with their product team and dictating just what the hell they were going to do—and motherfuckers, you’d better tell me what I want to hear or you’re off Facebook Island, not that you were even invited in the first place. That approach had more or less worked, particularly in the early days, as every outside partner jockeyed for the prestige of a place among the first or second wave.

Except for Amazon.

There, the always tenuous and yet strategic relationship overrode my PM droit du seigneur over FBX. That was the meeting my phone was reminding me of, and with a groan of exasperation I dropped everything for it.

To paraphrase Lord Palmerston about great nations, companies like Facebook had no eternal allies and no perpetual enemies, they merely had eternal and perpetual interests. Despite the various “partner programs” it ran, Facebook didn’t really have partner companies, much less real friends. It perceived the world as populated by either fearsome enemies that presented existential threats—and there were only a few of these—or companies that presented convenient temporary alliances. Facebook’s good graces were always conditional, and it was a foolish company indeed that took them for granted.

Amazon clearly fell into the former category.

Jeff Bezos was a maniacal leader who would stop at nothing until his vision of the world was realized, and who had inspired, cajoled, or intimidated an army to implement that vision. Zuck looked over at Bezos in Seattle, or Larry Page from Google in Mountain View, or (in the past) Steve Jobs from Apple in Cupertino, and he saw more than just a tech company and a chief executive. He saw a reflection of himself among those men, and that was terrifying. Other players in the tech or media worlds could be outwitted, out-engineered, or otherwise co-opted or bought off, but not these alpha companies. With Amazon or Google, the Facebook army had to close ranks and present a phalanx, and possibly battle an equal foe whose CEO was just as much of a kamikaze as Zuck was (Carthage must be destroyed!).

(Twitter didn’t even enter this worldview, and was merely a distraction. Zuck had once famously derided the company as “a clown car that drove into a gold mine,” and that’s possibly the last time anyone at Facebook had thought about it.)

And so any business-development process with those Great Power companies was always a grinding negotiation on the order of an arms-reduction treaty: lots of internal pomp and circumstance; interminable meetings held at long, crowded conference tables; and very little incremental progress. When a contract was finally produced, it was batted back and forth by the lawyers, and covered in red lines rejecting one or another clause.

With good reason too. These companies without question were smart, thought in decades rather than the years (maximum) of most tech companies, and had the wherewithal to project their own agenda despite Facebook’s wishes. Oh, and while we all pretended to be friends, nobody was even remotely stupid enough to believe that.

The meeting with Amazon went well, at least at the engineering and product level. The specific team involved was called A9, which was the Amazon subsidiary wholly responsible for that behemoth’s search and advertising technologies.* Effectively, it was the private, in-house technology shop that built Amazon’s own DSP, the brokerage house that let Amazon buy ads anywhere on the Internet to drive you back to your Amazon shopping cart. The company spent hundreds of millions of dollars on advertising, and it could turn on FBX revenue in a huge way for us. Naturally, A9 would be the group that would develop any FBX technology, as it had already done on Google’s exchange.

Amazon’s engineers were sharp, knowledgeable, and knew the programmatic ads-buying world inside and out. They got it, instantly, and I was looking forward to having them build their sophisticated tools on our newborn platform, and Facebook competing with Google for their inexhaustible ad budgets.

Then the business guys got involved.

A9 was run by Matt Battles, who through the course of this three-hour meeting (and others yet to come) revealed himself as a sort of Svengali character, grasping all the important technical and product details of our proposed collaboration, and adopting a pose of aloof and calculating skepticism about the whole thing. He seemed willfully tenuous about the proposed calculation, almost as if to reinforce that it meant more to Facebook than to Amazon to get this deal done (which it probably did).

Our Facebook business-development person, as well as the Amazon partner manager, reflected that poker-playing pose, each adopting his own corporate stance of scrutiny and skepticism.* There were lots of hard questions from both sides on how data would get used, and potentially recycled inside the obscure inner workings of the other company. Would Facebook record what Amazon ads were shown to which person, and use that data? Even in a distant, second-order way, would Amazon ad performance be used in some Facebook model around clickthrough rates? Facebook in turn wondered how far into the Amazon machine all this user-level impression data would flow, and if the product recommendations on amazon.com would suddenly reflect your interaction with ads on facebook.com. The two giants circled each other warily, neither willing to take its hand off its concealed weapon and shake hands instead.

And so the corporate standoff went . . . for months. Amazon would eventually join FBX, and its ad spend would benefit the always-demanding FBX revenue dashboard, but not until many meetings and contract negotiations later.

And what of that other stalwart Facebook competitor, Google?

The initial noises around Google Plus had been more than alarming, as it became increasingly clear that their foray into social media wasn’t some halfhearted effort to knock off a pesky upstart. As had slowly leaked out over the past year, either via the press or current Google employees, all of Google’s internal product teams were being reoriented in favor of Google Plus. Even Search, then and now the most frequented destination on the Web, was being dragged into the fray, and would supposedly sport social features. Search results would vary based on your connections via Google Plus, and anything you shared—photos, posts, even chats with Friends—would be used as part of Google’s ever-powerful and mysterious search algorithm.

This was shocking news, even more so to Googlers. Search was the company’s tabernacular product, the holy of holies, the one-line oracle of human knowledge that had replaced libraries and encyclopedias.

By all accounts (and Google information security was clearly not as good as Facebook’s) this caused a considerable stir internally. In January 2012, at a company-wide Q&A, Google’s founder Larry Page addressed this new direction forcefully, quelling the internal dissent and issuing a Googler ultimatum.

“This is the path we’re headed down—a single, unified, ‘beautiful’ product across everything. If you don’t get that, then you should probably work somewhere else.”

With the gauntlet thrown down, Google products were soon ranked via one unique metric—how much did they contribute to Google’s social vision?—and either consolidated or discarded appropriately.

As part of the budding media seduction around this new product, Google posted whopping usage numbers. By September 2012, Google announced the service had 400 million registered users and 100 million active ones. Facebook hadn’t even quite reached a billion users yet, and it had taken the company four years to reach the milestone—100 million users—that Google had reached in one. This caused something close to panic inside Facebook, but the reality on the battlefield was rather different from what Google was letting on.

This contest had so rattled the search giant, intoxicated as they were with unfamiliar existential anxiety about the threat that Facebook posed, that they abandoned their usual sober objectivity around engineering staples like data and began faking their usage numbers to impress the outside world, and (no doubt) to intimidate Facebook.

This was the classic new-product sham, the “fake it till you make it” of the unscrupulous startupista, meant to flatter both ego and chances of future (real) success by projecting an image of present (imagined) success.

The numbers were taken seriously initially—after all, it wasn’t absurd to think Google could drive usage quickly—but after a while, even the paranoid likes of FB insiders (not to mention the outside world) realized Google was juicing the numbers, like an Enron accountant would do a revenue report. Usage is always somewhat in the eye of the beholder, and Google was considering anyone a “user” who had ever so much as clicked on a Google Plus button as part of their usual Google experience. Given the overnight proliferation of Google Plus buttons all over the Google-user experience, like mushrooms on a shady knoll, one could claim “usage” when a Google user so much as checked email or uploaded a private photo. The reality was Google Plus users were rarely posting or engaging with posted content, and they certainly weren’t returning repeatedly, like the proverbial lab rat in the drug experiment hitting the lever for another drop of cocaine water (as they did on Facebook).

To further stoke the fighting spirit (as well as internal trollery), the face of Google Plus was a perfect target of Facebook-ian contempt. Vic Gundotra was a former Microsoft exec who’d climbed the treacherous corporate ladder there before jumping to Google. It was he who had whispered a litany of fear into Larry Page’s ear, who had greenlit the project, and it was he who headed the rushed and top-down effort (unusual for Google) to ship a product within an ambitious one hundred days.

The man oozed an off-putting smarminess as he stumped loudly for Google Plus in countless media interviews and Google-sponsored events. What was most insulting to a Facebooker was his studious omission of any mention of the social media behemoth in any public statement, as if the very raison d’être for his now-towering presence at Google didn’t even exist. Like some Orwellian copywriter, engineering language and perception to suit a nonexistent fictional reality, Google would never explicitly mention the Facebook elephant in the room in any public statement. “Networks are for networking,” intoned Gundotra. “Circles are for the right people,” he continued, referring to Google Circles, a way of organizing social contacts shamelessly copied from Facebook’s long-ignored Lists feature.

Within Facebook, Vic assumed the role of Emmanuel Goldstein from Orwell’s 1984, and many were the rips and jibes that he suffered in internal Facebook groups, a socially mediated Two Minutes’ Hate, whenever someone posted a link to one of his pro-Google bloviations. This had gone beyond mere corporate rivalry and had become a personal struggle to many Facebookers who saw their identities wrapped up in the company, Facebook as an expression of themselves (or was it vice versa?)

“Gokul, I’ve only got three people to create an entire platform and business, and produce tens of millions in revenue in a few months.”*

The scene was the final minutes of my more or less weekly one-on-one with Gokul, my official manager, although his management was largely symbolic in those chaotic days.

“Three, Antonio! It’s like the movie 300. You’re Spartans, man. Spartans! You can do it.”

Gokul was referring to the cinema version of the Frank Miller comic, which had come out a couple of years earlier, and had been a hit among the geek set. It depicted the early days of the Greco-Persian Wars of the fourth century BC, as filtered through modern superhero aesthetics.

“You realize, Gokul, that the Spartans lost at Thermopylae, right?”

He shook his head, looked away, and said nothing, as he always did when confronted with an awkward reality. We had reached his stand-up desk, on which stood an unused monitor (Gokul didn’t have the patience to plug in the cable). It was clear our weekly one-on-one was over.

I walked back to the FBX hooch with nothing to offer the team in terms of increased resources. Despite an auspicious beginning to FBX, and strong public and media interest in the product, Facebook was unwilling to invest anything beyond a couple of engineers and a handful of part-time support staff. This may seem incongruous, but it was in keeping with the Facebook zeitgeist. FBX, despite its initial promise and my vision of it as the centerpiece of Facebook’s future ads-targeting product, was still considered a rushed revenue stopgap, puking out cash the company needed to make post-IPO numbers. Never in over a year of lobbying did anyone one level above me in management truly understand, much less actively support, the real-time programmatic ads-buying technology that was then taking over so much of the advertising world. To the extent it could produce revenue right now! it received the lukewarm support of Gokul and Rabkin, and the tolerance of the sales side, which took its cues from product anyhow. My naive assumption was that if I could only make FBX into a billion-dollar business, I’d convince the recalcitrant (to not say willfully ignorant) Facebook management of the size of the overlooked opportunity, and the tide of opinion would shift.*

This was, of course, madness. Money talks, but inside Facebook at least, it spoke only if what it said was part of the bigger, accepted narrative.

As with the lawsuit during the AdGrok days, I shielded the engineers from all this political brouhaha, and acted as if it was a given that all these wonderful features we’d dreamed up—many of which we’d implemented by this point, resulting in a slew of patents—would one day see the light of day.

But in fact it was us against the world, and Facebook management.

To be someone or to do something, which would I choose?*

A man occasionally reaches a fork in life’s path.

One road leads to doing something, to making an impact on his organization and his world. To being true to his values and vision, and standing with the other men who’ve helped build that vision. He will have to trust himself when all men doubt him, and as a reward, he will have the scorn of his professional circle heaped on his head. He will not be favored by his superiors, nor win the polite praise of his conformist peers. But maybe, just maybe, he has the chance to be right, and create something of lasting value that will transcend the consensus mediocrity inherent in any organization, even supposedly disruptive ones.

The other road leads to being someone. He will receive the plum products, the facile praise afforded to the organization man who checks off the canonical list of petty virtues that define moral worth in his world. He will receive the applause of his peers, though it will be striking how rarely that traffic in official praise leads to actual products anybody remembers, much less advances the overall cause of the organization.

I certainly had options. I can’t say I didn’t.

I had outs, and could have walked away from FBX.

All I had to do was shut up, keep my head down, go through the motions with whatever shitty product was handed to me after FBX, and read along from the Facebook script about how to be a proper product manager. I might even resurrect my career if one of those products was “successful.” In the rapidly ossifying Facebook culture, one could get along simply by going along.

Plus, there were real stakes on the line.

Consider the number of shares and the fact that at the time of this writing, in December 2015, Facebook is worth almost four times what it was when I struck my AdGrok deal. I had two children, whose mother’s nine-to-five corporate job provided for them, though just barely. Her London trader heyday was long gone, and now her salary was that of your standard-issue MBA inside the corporate machine. I contributed a slab of cash every month, per California support guidelines, but it would be my Facebook stock vesting yet to come that would pay for private high schools and Stanford, so Zoë and Noah wouldn’t have to sneak into this country’s elite through the back door from cattle class, as I had to do.

Short of me pulling another startup out of my hat, any windfall would come only from the Facebook shares I was so cavalierly risking in my mad pursuit of FBX. But the ego drives, and honor compels, and stubbornness steels the will that moderation and judgment might sap. Faced as I was with an array of opponents, my reflex was the same as always: double down and dig in.

Like Col. Boyd, I’d choose to do something, rather than be somebody, and tie myself to the mast of the FBX ship, wherever it took me and whatever the future cost.

JANUARY 16, 2013

This was an embarrassment.

The senior leadership of Omnicom, one of the largest ad agencies in the world, was sitting on the other side of a long conference-room table in Building 17.

My side of the conference table was populated by the Ads senior leadership, who came out en masse only for such big-ticket guests, and the line PMs and engineers involved in the newfangled FBX and Custom Audiences stuff that had gotten so much outside attention (Mathew Varghese, me, the odd engineer).

There were four ad agencies, so-called holding companies, which served as umbrella organizations for the hundreds of smaller agencies that ruled the media world. Collectively, their revenues were in the tens of billions, and the total media spend they controlled or influenced was many times that. These motherfuckers were the real deal, the absolute A-plus, Big Swinging Media Dicks. Sitting like Zeus on Olympus, they could fire lightning bolts of media spend hither and yon, activating any publisher or media channel instantly. Facebook had been patiently seducing the agencies for years, with entire teams in its New York office dedicated to teasing spend into our new, unproved medium.

For years, that pitch had focused on the ephemeral promise of Facebook pages and Likes. But with Custom Audiences and Facebook Exchange, the agencies’ ears had perked up. Suddenly, they could actually use their data and targeting on Facebook in ways they could account for, proving worth to their clients, which would generate more spend, and set the media wheel giddily turning.

At the head of the table was Josh Jacobs, the head of Accuen, Omnicom’s “trading desk.” That’s right, with the rise of programmatic media and its exchange-like real-time markets, even the stodgy old agencies were opening business units that operated like those on Wall Street. These were the brains of the agency world—to the extent that world possessed any—and they were very savvy and bullshit immune.

“So when can we get the identity-matching stuff in via FBX?” asked one of Jacobs’s lieutenants to his right.

This guy was actually in a suit, about as rare as a wallaby on Facebook’s campus. He was the head of some subbranch of the Omnicom empire that ran retargeting for the holding company’s clients.

I sure wasn’t going to answer for the current internal shit show that had haunted the Facebook side of the table for months now. Gokul, as the senior product person present, took the question.

“We’re having lots of internal discussions about that right now, but we basically haven’t quite come down to a company-wide decision on our future direction.”

Ugh.

What a cop-out. Burn my Facebook fleece now.

An awkward pause followed the obviously unsatisfactory answer. They sat there with their tailored shirts, good shoes, and chunky power watches, looking the very picture of East Coast Media Douchebags. They stared at our side of the table, where we slouched in our hoodies and sneakers, and waited for answers.

But there were no answers to be had.

For the past six months a vicious debate had taken place internally at Facebook about the directions of CA versus FBX. The crux of that debate, which I touched upon earlier, boiled down to the following: As Facebook waded into the world of outside data—desktop browsing, mobile app installs, offline purchases in stores, and so forth—would it build an open system that allowed advertisers to bring their data to Facebook, while preserving Facebook’s user privacy and preventing data leakage to the outside world? Or, as it had in the past, would it attempt to capture as much of the ads management stack as it could, even if it meant shipping inferior product, shunting out any and all companies other than the end advertiser?* The implications for revenue, arguably the most important issue in the debate, were severe.

The completely “closed” stack that Boland was backing, which would involve an expanded version of the Custom Audiences product, would take years to build. Also, it would take years to convince advertisers to use it, as they generally preferred independent tools for ads serving and delivery that weren’t so reliant on a single publisher like Facebook. Also, inertia alone would disincline them from using it, as the real-time technology that was then taking over online media, and of which FBX was an expression, was fast becoming the dominant paradigm for how sophisticated ads targeting was done.

What Omnicom and its deep-pocketed advertisers wanted was what they called “data portability”; that is, the ability to use their valuable first-party data, or expensively acquired third-party data, wherever it wanted to, whether on Facebook or in the New York Times. By building some tottering edifice of technology, and then limiting access to Facebook’s inventory through that technology, Facebook was effectively maintaining the walled garden it had tended for years. It would brick over the narrow opening in that wall we had created with Facebook Exchange in the first place, and which had created the excitement that had led to this meeting.

The only way Facebook would convince the Omnicoms of the world—and by extension, every sophisticated advertiser—to adopt Boland’s closed-stack direction was by driving so much more value via the use of Facebook’s identity and cross-device data that the advertisers would swallow the bitter pill of Facebook’s hard-to-use technology. As the value of that unique Facebook identity was still not completely known, this represented a risk.

The alternative was the “open” stack that the FBX team (by which I mean me, and just about only me) was loudly advocating. By building technology consistent with the growing standard, we would make adoption by advertisers quasi-instantaneous. The FBX team had already semisecretly built (but not shipped) the basic technologies that would have safely exposed Facebook identity across browsers and mobile devices. It was merely a matter of turning it on, something I had been harassing Facebook management to do for months.

To cite our perhaps overused financial analogy, imagine Facebook was a large financial-services company like Fidelity. What Facebook was proposing in the “closed” plan was this: Facebook didn’t just want to offer its unique mutual fund, stock, and bond investments; it also wanted to run the exchange where those investments were traded, the brokers buying and selling on that exchange, and the stock-trading tool used to log in and trade, as well as the entire financial system around the investment platform—namely, banks, checking accounts, credit cards, the works. If you wanted to buy into the Facebook investment fund, everything from your debit card to your 401(k) retirement account would be labeled “Facebook,” and work only with other Facebook-labeled products. Facebook would exist as its own financial world, completely cut off from whatever Schwab or Bank of America world you came from.

What if you wanted to buy something other than a Facebook mutual fund in this Facebook-owned and -operated financial empire? Facebook would eventually allow you to do that as well, but only through its exchange and buying tools. So you could buy a few shares of Google or General Motors, but you’d still log into the Facebook buying platform.

Would this work?

Well, maybe. If Facebook’s investment offerings were so uniquely high-performing, and if your ability to buy outside stocks and bonds was actually amplified by doing it via Facebook’s platform (versus via all the other means for buying those commodity products), then yes, people would put up with the hassle of moving their entire financial lives into Facebook-land. Not to mention, they’d put up with the very real risk of lock-in such a move represented. (Once all your financial details and data were inside Facebook, your ability to unplug from the beast and take your business elsewhere would be severely limited.)

On the technical side, the decision had major implications as well. Continuing our analogy, Facebook was technologically like some local credit union in Omaha, Nebraska, though with the deposit balances to rival Bank of America. It would have to offer every banking service known to man if it wanted to compete with the entire financial world, and that would take time, even with Facebook’s inhumanly fast engineering.

Revenuewise, that meant the monetary value of this whole identity vision wouldn’t be realized for years either. Such a construct, given that it was much more than the sum of its parts, would assume its real value only when it all existed. What value did Custom Audiences have if you could use its wonderful, person-centric targeting only on Facebook and not on the rest of the Internet too? If Facebook had an ad server (and we’ll get into what that means in a moment) and it worked only on Facebook, who cared? An advertiser still needed its old ad server to show ads on Google-controlled Web properties and everywhere else. This glittering vision in Facebook Blue of everything running on Facebook would take years to realize.

By comparison, the value of the open plan would be realized instantaneously. If FBX had access to the identity matching we’d built on the Custom Audiences side, not to mention ads inventory on mobile, then the exchange could serve as a central clearinghouse for online identity everywhere. Everything we were ideating around identity, that entire spiel I regaled you with about names, both online and off, could be realized in a few weeks, all due to the open and standardized nature of FBX. And all of it implemented with three engineers and one cocky product manager, and not the roomful of bickering, big-company clowns that every meeting on the matter tended to feature.

Of course, to be fair to the closed plan, if it succeeded and Facebook somehow convinced the world to leave the outside Internet behind and jump headfirst into FB land, it would leave Facebook in a magnificently dominant position overall. That was the real question behind this debate: Was the performance of Facebook’s ads and identity data enough to convince the world to abandon everything else for Facebook? If not, then the closed plan was Sponsored Stories Part II, and the flying saucers would again fail to appear.

This closed-versus-open debate would consume rivers of electrons in acrimonious email debates, during which your humble correspondent used potentially offensive language to describe otherwise respected members of the Ads leadership, sowing seeds that he would exponentially reap in the sweet by and by.* But right then, those email debates were being batted brusquely back and forth, and sitting in the very inboxes we all discreetly checked as the Omnicom meeting droned ever on.

Though not in that room with Omnicom, a new presence was making itself felt on the Ads team right then, a presence that would change everything for Facebook Ads. Before I introduce that transformative character, there’s something critical about Facebook’s culture, which we’ve only brushed against thus far, that you need to understand:

Facebook was an empire Napoleonic in both tenor and flavor, and like any hive of palace politics, one’s proximity to the emperor around which the pageant swirled was a direct function of his or her institutional stature. Get pulled into the orbit of the Boy Dynast of Facebook, and your product got a kick of institutional momentum via the implied royal blessing: engineers would be allocated, other products canceled or pushed out of the way, and, most important, you became a fixture in the courtly firmament of the empire. Even if your product failed (and many were the stillborn or grotesquely misshapen offspring of Facebook’s anointed inner circle), you lived to ship and influence another day, star undiminished, reputation unmarred. Several were the princes of Zuck’s court: Blake Ross, Justin Shaffer, Sam Lessin, and many more whom a commoner such as I, rusticating in the Ads wilderness, could observe only from afar.

Fortunately, Facebook itself served as the society pages to its own courtly intrigue: the quasimaniacal Facebook usage among employees meant every debutante’s coming out with Zuck, or every kingly banquet at In-N-Out Burger, was there for all to see. Like a Brit on a bar stool scanning tabloids for news of the royal family, the Facebook hoi polloi friended or followed whomever they could among the Facebook nobility, and watched, nose pressed against the glass, the regal extravaganza unfold.

One of those princes, first among equals, was Andrew “Boz” Bosworth, Zuck’s former teaching assistant at Harvard. Purportedly raised on a horse ranch in Northern California, he was a big, bald bully with tattoos on both arms, a former member of the Harvard tae kwon do team grown thick with age. Prior to his involvement in Ads, he was the engineering brains behind the initial versions of News Feed, a fact he’d never fail to mention on a first meeting. He was the first Friend of Zuck to work in Ads, and it was understood he had been dispatched to unfuck the Ads team from the maelstrom of directionless chaos under Badros and Gokul.

In that sense, Boz was an absolute godsend, and the Ads team, not to mention Facebook’s finances, has reaped the benefit of his decisive authority ever since he assumed the reins. As I hope I’ve managed to detail, from 2011 until early 2013, the Facebook monetization strategy distinguished itself mostly by its complete absence (other than the failed Sponsored Stories experiment). As with the British schoolchildren in William Golding’s Lord of the Flies, when product managers and engineers were left to fend for themselves in a state of anarchy, it wasn’t long before a boulder got dropped on the fat weakling, and there was a rotting pig’s head on gleeful display. Boz was like the British naval officer appearing in the book’s murderous climax: the feral boys, on the cusp of truly losing it, awaken, and decide to take up civilization again.

Which is more or less what happened in early 2013 when Boz started appearing in Ads meetings, and eventually took over. Suddenly, Ads could make decisions without resorting to Sheriff Sheryl, since we had our own viceroy, our own official Vicar of Zuck, running the Ads show.

But there was downside to aristocracy.

Almost comically, and against all Facebook tradition, Boz named his personal conference room after himself: LiveBoz. Decked out like a personal man cave with couches and low-slung coffee tables, rather than the upright Aeron chairs and long tables of most conference rooms, it was where we had the penultimate and testiest debate about product direction. He also seemed to have a team of professional photographers following him around in his personal life, the resulting flurry of paparazzi photos peppering his Facebook Timeline, which resembled something between an ever-present wedding album and a running Hollywood movie premiere (but with few celebrities other than himself).

In Katherine Losse’s so-so firsthand account of life at early Facebook, The Boy Kings, she describes a mafia of Harvard recruits. They were core hires, and would go on to form the backbone of future Facebook. By all reports, the atmosphere at early Facebook was bro-y and redolent with the joshing violence typical of young, hormonal males. One of the Harvard mafiosi got into the habit of threatening other engineers that he’d “punch them in the face” if they messed up. That person was Boz, as I was able to guess while discussing the matter with a Facebook old-timer who dished details.

Quod licet Bozi, non licet bovi. Gods may do what cattle may not.

Boz might have been a god at Facebook, but what he wasn’t was an ads man, as he knew nothing about the monetization space. This was no ding; most of Facebook, outside of the Ads team, had no clue how Facebook made money and the free food and shuttles got paid for. Most were quite content to stay that way, and blissfully outsourced this concern to the Ads teams and Sheryl and kept on working on whatever piece of the user-facing product was theirs.

As a form of rapid on-boarding, Boz was invited along to client meetings with the product marketers and managers. On the one hand, this was slightly preposterous. He had zero qualifications, other than Zuck’s confidence, for the job.

On the other hand, who are we kidding? It was advertising technology, not orbital mechanics; a mentally agile techie, having spent an action-packed career learning (if not creating) new technologies and business models, could learn enough high-level basics to make (mostly) intelligent decisions—particularly if the entire team bent over backward to train you in a hurry. That’s effectively what happened, and from late 2012 on, Boz was increasingly involved in Ads strategizing and general mass-email-thread politicking. While silent at first, getting his bearings in the new subject matter and political landscape, he’d soon grow very vocal indeed.

But before that happened, Facebook Ads, thanks to Boland no less, had one of those sudden, out-of-left-field, WTF moments that changed everything.