Ad Majorem Facebook Gloriam*

There is a tide in the affairs of men,

Which, taken at the flood, leads on to fortune;

Omitted, all the voyage of their life

Is bound in shallows and in miseries.

On such a full sea are we now afloat,

And we must take the current when it serves,

Or lose our ventures.

—William Shakespeare, Julius Caesar

MARCH 18, 2013

Boz was no longer the dilettante guest of the Ads team, sitting in like a tourist on meetings, but the officially installed leader, Badros having been given the bump. He carried himself like he owned the place (which, of course, he kind of did). Boz’s takeover of Ads represented more than a cosmetic shift in the organizational chart; we were witnessing a complete reconfiguration of the Ads space-time continuum along a single dimension from “Boz likes you” to “fired.” Where you were on that continuum was yet to be determined for most Ads folks, and I had no idea where I fell.

Presumably to figure this out, Boz requested we have a one-on-one, even though I didn’t report to him officially. In short order, I found myself sitting across from him at a small table in a small conference room, staring at the tattoos gracing his forearms. On his right there was a figurative map of California, on his left the word “veritas” encircled his wrist (“truth” in Latin, also Harvard’s motto).

I had, of course, done a complete Facebook stalk going back through his time at Harvard and clear into babyhood. Thanks to the FB hashtag #tbt, I had seen his idealized and mediated autobiography from rural Northern California to Harvard to Facebook.* I had also scoped out his management philosophy, or really, his public conception of it. He clearly was one of those tough but fair types who pride themselves on directness and honesty. Truth was written on his very body, after all.

For my part, outside of physics textbooks, I found Truth to be a rather rare commodity, particularly in the tech world. I had also noticed that those who most made a big show of believing in Truth were unusually attached to whatever well-groomed pack of lies they held dear.

The conversation proceeded quickly to my role in the Ads team. At Facebook, the performance review cycle was semiannual, with every February and August bringing the usual political jockeying to make sure your biggest fans reviewed you, and your biggest detractors didn’t quite manage to.

“You’re a very divisive figure, Antonio. I read the reviews of your team members, and then that of management. They’re completely opposed. One loves you, and the other hates you.”

Yeah, no shit.

I could imagine the feedback, without having been told the specific authors: Reesman, Hari, and Gary and others on the FBX team all gushing praise for my dedication and product leadership. And then Gokul and Boland decrying my uncouth insubordination, smart-alecky arrogance, and corrosive criticism of current Facebook strategy.

“That’s true, Boz. I’ve certainly made friends and enemies here. But my goal has always been to give Facebook the best ads system possible.”

This was, in all seriousness, true. I could barely remember what my life was like before Facebook, and there was a trail of destruction I had caused by spending my entire life there: two children neglected, two different women whose worthy love I’d spurned, two boats rotting in neglect, and anything like an intellect or a life outside campus nonexistent due to indifference and my dedication to the Facebook cause.* Don’t be deceived by my withering treatment of Facebook in this book; inside every cynic lives a heartbroken idealist. If I’m now a mordant critic, it’s because at one point, like Lucifer once being the proudest angel before the fall, I too lived and breathed for Facebook, perhaps even more than most.

We moved on to the topic of the open versus closed debate that was roiling all of us. While he was equivocal on the matter, he expressed his firm intent to come to a conclusion quickly and end the uncertainty that plagued every forward-looking decision Ads was making. This was one of those Delphic portents of old, which could be read either somewhat positively or negatively, depending.

Afterward, we stood up and shook hands over the small desk. The meeting had had the smell of a last meeting with the mafia capo before things turned ugly, and he started either unleashing hitmen or making offers that couldn’t be refused. I didn’t like the smell of it, not one bit.

APRIL 4, 2013

Boz, live. Again. This time in LiveBoz.

Gathered was everyone with a speaking part in the Great Facebook Debate of 2013: Boland, Gokul, Varghese, Boz, Scott Shapiro, and me. The stated purpose was to finally (finally!) hash out a final recommendation from Boz to Sheryl. Despite all the ballyhooed “final” meetings with Sheryl we’d had over the previous few months, we hadn’t reached a conclusion.

Given the bro-y décor of the Boz man cave, it was hardly surprising the meeting soon devolved into a male pissing match.

“Custom Audiences is way ahead of FBX these days,” Boland began.

Yeah, no shit, jackass. It’s me and Scott and Hari against you and the entire Facebook sales and operations organization you manage or influence, as well as the engineering teams running the targeting system and the API, all of which support Custom Audiences, while I have one engineer. It’s a miracle FBX is still even in the running.

My resentment at the unfair match aside, Boland was actually wrong, and for subtle reasons that he likely didn’t understand.

“That’s only if you count Custom Audience dollars as actually incremental. I think that’s a stretch to say the least. What’s in your Custom Audience revenue dashboard isn’t real revenue. They should probably be discounted by half, if we’re doing comparisons,” I countered.

Boland and I locked hostile eyes, and Boz interrupted in his managerial, peacekeeper role with conciliatory remarks.

Here’s where we really were:

For the past few months, since their respective launches in midsummer, FBX’s and CA’s revenues had both increased apace and were initially neck and neck. Since about December, though, and on paper (or in Facebook’s revenue dashboards, which were what mattered) CA’s revenue had edged out FBX’s revenue significantly. The reasons for that were telling.

As mentioned earlier, FBX’s technical challenges were far, far greater than CA’s. We were creating an entirely new ads infrastructure, while CA was merely coasting on existing Facebook targeting technology, recycling much of the targeting logic that already existed. I would know; it’s what I’d spent the previous year building. Clunky though its technology was, it marked an expansion of basic functionality already inside Facebook’s API and interfaces, which meant existing Facebook advertisers could easily start spending money on it. They could do so directly via Facebook’s ads-buying interfaces (for which there was no FBX equivalent) or via the third-party ads-buying tools of Facebook’s advertising partners whom Facebook had pressured into supporting CA features.

FBX was designed to work seamlessly with the outside ads world more accustomed to real-time exchanges, but it wasn’t without its technical and adoption issues. As I’ve explained, FBX required deviations from the standard programmatic playbook, such as different ad formats. This required many epic hacks by the FBX team to try to wedge an entire ads system that had evolved separately into the industry standard of online retargeting. Several aggravating months and a few patents filed later, we had largely succeeded, but that convergent iteration around market needs was something for which Facebook culture had very little patience.

Also, you had to understand the psychology of the advertisers and their simplistic categorizations of the world. Then, and even now, advertising budgets were allocated by channel—TV, radio, Facebook, Google, billboards, whatever—almost like earmarked government handouts for one or another constituency. While the type of advertising—that is, retargeting people shopping online—was precisely the same (at the abstract level) via CA or FBX, spending on CA still fell under the “Facebook” budget; hence, whatever agency was handling the budget could easily steer spend toward it without much fuss, both businesswise and technically.

In FBX land, it was a different story: we were competing for spend with budgets that had been earmarked mostly for Google and its real-time ad exchange AdX, which was how most retargeting was actually being done. This was good, in that FBX spend was truly “incremental” (to use Facebookese); that is, FBX revenue was new dollars that were not otherwise being spent on Facebook. The fact these dollars were formerly being spent on much-hated Google made them even yummier, of course. Incremental spend—new money!—was the holy grail for Facebook at that point. Incrementality was, in fact, the challenge of any new ads product at Facebook. Sure, your dashboard showed $X million in spend, but how did you know that money wouldn’t have been spent on Facebook anyhow? There was no point in launching new products only to divvy up the same spend among a new set of buckets. This is why Boland’s numbers were mostly bogus revenue. At most fifty cents on the dollar was actually new money, the rest coming from budgets that would have been spent on Facebook in any case.

The 100 pecent incremental nature of FBX spend was a selling point to management, but, as mentioned previously, it also meant we had to convince Zappos or whomever, either directly or via our proxies the FBX partners, to explicitly allocate spend on Facebook instead of on other real-time channels. Then there was the additional fact that the advertiser had to design new ad creative to run those campaigns, which meant extra work for an as yet unproved channel. All retailers in existence had been burned spending money on Facebook before, and had fresh memories of their respective Facebook fiascoes. Since most retailers lived and died by their fourth-quarter performance around the Christmas shopping season, they were disinclined to experiment with anything truly new. Accordingly, CA felt more of the big Q4 spending bump than FBX did, with the net result that FBX had fallen behind. This weakened my hand in the ongoing poker game taking place in LiveBoz.

If I only had $1 million in spend per day to shut Boland up with, this meeting would be over very quickly, I thought.

But I didn’t. There was hope, though. FBX had recovered quickly given new features we’d shipped in 2013 to further ease adoption, and strong growth was back for Q1. CA, meanwhile, had started to languish again after the strong Q4. Boland’s team hadn’t shipped much of anything new beyond back-end tweaks that made the system less slow. Merely not totally sucking was not a very convincing sales pitch.

After the initial mutual broadsides, the meeting quickly devolved into the usual inconclusive corporate ritual, with Boz reiterating his mantra that we’d come to a conclusion soon. He vowed to consult each and every one of us sitting here, and then to formulate a formal recommendation to Sheryl. Sheryl would either make a decision or let him make it.

Pause and think for a moment.

What’s it like to be a Gokul or a Boland or a Fischer?

You’ve seen excellent engineers and product people come and go, as your recruiters land the best new grads from every leading school. Your privileged position inside a powerful and market-leading organization exposes you to every industry trend and player, and pads your network with a roster of power and influence. Your ability to make any company’s senior management appear and deliver fawning and revealing product demos means you know everything that’s going on in your industry, down to the color of the clickable buttons in the products.

And yet there you are, lips bolted like pipe flanges onto some bigwig’s ass, plodding along like an ox yoked to a grain mill, grinding the grist the big-company machine requires.

Why is that?

I’ll tell you why.

It’s because you are without a doubt the least daring and least innovative person at your organization, because in the opportunity-rich environment in which you live, the ambitious and capable have left to pursue it. There’s a negative selection in which the cream (or whatever it is that initially rises) gets constantly skimmed off, and you are what’s left after years of continual skimming. Changing from big company A to big company B is cosmetic, as it’s of course at least a lateral move if not a step up. You learn that what matters in a big company is to avoid falling victim to firing or layoffs, and to appear important and critical to the company’s mission. You have mastered the art of “managing up”: namely, controlling the feelings and perceptions of the management layer above you. You take feedback well, and make sure to be seen speedily acting on that feedback. If you have reports, you champion their careers internally (make sure they know you’re doing that), and try to mold them into people like yourself, who are organizationally effective and recognized as such. In all but the most pathological organizations, your reports’ success will reflect well on you and create your own success. You make sure to form allegiances and friendships with your peer managers, particularly in organizations like sales or business development that you’ll need to push your business agenda forward. When there’s an ineffective and incompetent member in the organization, rather than calling them an idiot to their face and firing them if possible, you channel feedback to their manager and learn to work around their incompetence. If the incompetence does not directly affect you or your team, you look the other way and focus on the levers you do control.

You’re middle management: you’re the necessary layer between the visionaries and risk-takers who created the organization, and the new acolytes of your religion for whom this is a job, and you are their first whiff of corporate culture and authority.

If you’re cleverer than most middle managers (e.g., Gokul), you’ll work at building your personal brand in a way that both augments your prestige and reflects well on the organization, all in a studiously self-effacing way that allays any concerns around thinking yourself a “star.” Failing that, the logo on your business card is your strongest asset, and you need to bank as much on that as you can, right up to the moment you trade it for another (hopefully better) one.*

These were the ingredients of the toxic meeting cocktail that was being poured right there in LiveBoz: the corporate aristocrat of Boz, the middle management lifer of Boland, and the obnoxious, shit-stirring self-assurance of the acqui-hired startup founder. It went down about as well as it sounds.

APRIL 12, 2013

The scene was the Only Good News conference room.

We’ve returned to our starting point, but under very different circumstances. The cast was almost exactly the same: Gokul, Boland, Rabkin, Sheryl, and now Boz. I had played every last card. FBX development had been pushed to its absolute max, despite what little resources we had. Using either flattery or deception, we’d gotten FBX partners to spend as much money as possible on the new platform. I had tried to charm and persuade, with whatever persuasive charisma had worked on investors and cofounders in the past, the other members of the Ads team to support a vision most of them barely understood.

That was the good.

On the bad side, I had been insubordinate to Gokul, more or less refusing to work on the bullshit make-work projects he’d assigned to distract me from FBX. I had been an insufferably obnoxious punk on the Ads team, pushing for a contrarian agenda in a culture rapidly losing its supposed tolerance for heterodoxy.

While FBX had been a qualified success, and I had received praise for doing much with very little, none of the erstwhile FBX fans in that room were now willing to stake any of their internal social capital on its future, much less on the overarching programmatic direction it represented. If Sheryl agreed to extend to FBX the data joining that was currently limited to Custom Audiences, or even to put mobile ads inventory (a product that would later dominate Facebook monetization) on FBX, there was a chance of pulling this off. Assuming FBX continued to grow at its current healthy rate, it would eventually dwarf Custom Audiences spend, in both incremental reality and what appeared in the dashboards. If Sheryl didn’t agree, then it would mean the death of FBX and everything around it: the technology itself, the innovative IP we had patented, the budgets we’d secured, the work the FBX partners had done to integrate, the bigger vision of it all. Gone. We were all-in and betting on just one card being drawn.

“You want to go ahead, Boz?” Sheryl gestured to Boz, who was sitting across from me.

“After talking to everyone concerned, including the Custom Audience and FBX teams, I think we shouldn’t ship identity matching on FBX, and continue with it in Custom Audiences only. That’s my recommendation.”

This was going to be a short meeting.

“Everyone here has contributed to this discussion, right?” Sheryl looked around the table. We all nodded.

“Well, if that’s what you think, Boz, then that’s what we’ll do.”

I caught Boz’s eye; he looked away.

Nobody moved or spoke. After so many months of discussion, it was hard to believe that a decision had actually been made.

Sheryl added finally, as if trying to snap us out of a stupor: “So that’s it. Nothing of Custom Audiences on FBX. No mobile inventory on FBX either. We’ll leave it as is, and keep pressing on with identity on Custom Audiences alone.”

I looked out the window at an alleyway that led to the main courtyard, and the huge HACK sign pointing skyward, like a divine commandment.

Well, I had certainly HACKed.

After some closing formalities that I was too distracted to register, everyone got up to leave. There was nothing left to discuss. Gokul glanced in my direction, but looked away the moment our eyes met, and darted out of the room. He was out of the building before I could even attempt to catch up.

I wandered into the FBX area, or what was left of it.

The product I had gambled my entire Facebook career on was now effectively on life support.

For the second time in two years, I walked out of the office in the middle of the afternoon with nothing to do.