Lean In: Witness Work and the Will to Monopoly Power

On Day 3, cross and re-direct of Mark Zuckerberg concluded, and Sheryl Sandberg shivved the FTC with kindness.

If yesterday aired out the hot docs in the FTC’s case to break up Meta, today threw some cold water on it. Meta took its turn crossing Mark Zuckerberg, scoring some points it wanted to land, and then the FTC tried to do cleanup, calling Sheryl Sandberg and using some of the same emails we’d seen with her old boss.

Inside the Aquarium

Mark Hansen, Meta’s lead lawyer, began with a module designed to show that Facebook had a laundry list of competitors. TikTok came first. Recall that the FTC puts TikTok outside the personal social networking services (“PSN”) market, but that including TikTok would dramatically reduce Meta’s alleged market share, undercutting the case for monopoly power. Hugging some language the FTC uses for its market definition, Zuckerberg testified that TikTok, like the Meta family of apps, has a “social graph” built using followers and contact importing.

Although I don’t recall Zuckerberg using the word “significant” in a prior answer (and it’s possible that he did at some point and I missed it), Hansen used that word in his next question: “What did you do as CEO of Meta when you realized TikTok was a significant competitor?” Zuckerberg answered that Meta ended up building Reels in response to TikTok's “competitive threat,” which was a “meaningful headwind” for the business. Meta made a large investment in what it calls “MRS”—Modern Recommendation System—an AI content recommendation algorithm. This was such a company priority that progress on MRS came up in board meetings and weekly meetings of senior executives, Zuckerberg said. An email from a Meta exec flashed on the screen, reading: "TikTok's growth so far is worrying and their time-trend projections have them overtaking Instagram in the US in terms of total Time Spent." It was an effective sequence.

Next up was YouTube. Zuckerberg testified that he’s seen recent statistics showing that users spend more time on YouTube than on Facebook and Instagram combined. In a mismatch, Hansen asked, “When did YouTube emerge as a competitor,” and Zuckerberg answered in terms of when YouTube launched—Zuckerberg testified that a YouTube founder was an early engineer at Meta, likely referring to Steve Chen, who was at Facebook for a few months. Of course, the Q&A setup here made it seem like YouTube was always a competitor of Facebook’s even before Facebook had its video features. A sleight of hand. So too with the next question about iMessage. Hansen called it a “leading competitor” before asking if it had a social graph. At some point, there has to be an objection that the lawyer should not be testifying. Chief Judge Boasberg, though, will see through it.

Then Hansen shifted to asking about former competitors. Without the help of counsel to lead him, Zuckerberg gave an answer that actually made MySpace, Twitter, and YouTube look like they weren’t competitors—those platforms focused on public sharing and “going viral,” while Facebook’s “unique hook” was private sharing, i.e., with family and friends.

Credit: Tom Anderson, founder of MySpace

Hansen shifted to making a relatively straightforward point that network effects weren’t enough to save MySpace. “Network effects” is an economic term that describes a product that becomes more or less valuable as it gains or loses users. Usually it’s used to describe a positive feedback loop: more people join as more of their contacts are on the platform, which leads more of their contacts to join. Facebook executives and Zuckerberg in particular used the term a lot in their communications. But Zuckerberg took the MySpace question as an invitation to give a Ted Talk on what he calls “negative network effects”—not wanting to post things to Facebook because you’ve accumulated too many connections over the years that you don’t want to unfriend because of social pressure. An AI “discovery engine” can filter these connections down to the people you really want to see. It was an interesting point in a long digression, made funnier when Hansen ignored it and went right back to asking about MySpace in 2006. Zuckerberg gave his view that MySpace was “chaotic”: users could pick their profile background and autoplay music on the page. Facebook, in his words, had “a cleaner, faster interface” that was more appealing to users.

Chief Judge Boasberg, for his part, interrupted asking if network effects are less important now that users can share outside or across networks. Zuckerberg said yes, pointing to contact importing and the nature of content going viral, like the recent OpenAI Studio Ghibli-style memes. A follow-up from the court: “How much does it matter if friends are on a particular platform if friends can share outside of it?” Zuckerberg called it a “good question” (one of his favorite Zuck-isms) and explained that it doesn’t matter much because people can fluidly share across platforms, using each one for its value as a “discovery engine.” He'll choose whatever the right medium is—short-form video, photo, story, text—for the moment, moving between them. That supports the concept of “multi-homing,” an economic term introduced in Meta’s opening that we’ll cover when we get to Meta’s economist.

For a final former competitor, Hansen turned to the Path app, using an email the FTC introduced into evidence yesterday, where Zuckerberg wrote:

"Theoretically we could survive . . . Instagram, etc growing quite a bit, but if Path grows and isn't deeply wired into Facebook then that would be a big problem for us."

We learned some of the history of Path, founded by a former Facebook engineer. And we learned what the “Aquarium” meant in Zuckerberg communications. That was Zuckerberg’s term for his all-glass conference room, which over time became synonymous with the leadership team that met there. He has an all-glass conference room, he told us, because it’s “good to have leadership transparency.” That segued into a discussion about how Zuckerberg likes to shoot from the hip with long emails to work out ideas in the interest of transparency, designed to pull the sting from the language we saw in emails yesterday. “I definitely wouldn’t characterize myself as someone who holds my cards close to my chest,” he said—actually an unhelpful admission for Meta about why the court should take his emails seriously. He likes “entertaining wild ideas that are low probability.” More than once in the past few days, he invoked former Intel CEO Andy Grove’s line that “only the paranoid survive.” Zuckerberg summed up: “It’s important to think about things that might seem close to unimaginable because sometimes those things happen. That’s just my style.” Writing is, for him, “a way to flesh out the ideas in a more structured way.”

In any event, explaining why Facebook was giving promotional help to Instagram but not Google+ or Path, Zuckerberg said that he saw Path as a threat but “viewed Instagram as adjacent to what we were doing.” For my money, Chief Judge Boasberg seemed visibly puzzled by that answer, as it was so clearly contrary to every other statement we read from Zuckerberg over the last two days’ worth of emails.

Tech Optimism

Back to Instagram. We looked at an email from Zuckerberg to Instagram co-founder Kevin Systrom promising the moon if Instagram joined Facebook. Zuckerberg testified that after the acquisition closed, Instagram “plugged in” to Facebook software infrastructure “to improve critical parts of the service”—helping to scale by improving recommendations on who to follow and detecting spam. Asked if he was worried about Instagram taking usage away from Facebook, Zuckerberg said no—those concerns came later once “Instagram got to a massive scale and we were operating a much different company”. Responding to the FTC’s exhibit yesterday showing that Meta admitted Instagram would have 300-400 million users even without Meta, Hansen asked if all the good things that happened to Instagram would have happened anyway. Another speech followed, the point of which was that scaling to 1 billion users is hard without being acquired by a parent company—”not impossible . . . . but does it seem like it would be likely? I think extremely not.” Only Facebook did it alone, clearly a point of pride.

But Zuckerberg’s chutzpah as a witness was growing. Before moving to a module about WhatsApp, Zuckerberg gave another speech about how “AI and glasses”—probably referring to his own Meta Glasses product—“are the next major computing platform” and that it’s “unthinkable” to him that we won’t be wearing smart glasses in the future. Google co-founder Sergey Brin, who spearheaded the much-maligned Google Glass, is probably licking his chops that Zuckerberg successfully marketed what he thought of first.

The cross wrapped up with WhatsApp. Hansen showed Zuckerberg his email summary notes from a meeting with WhatsApp co-founder Jan Koum circa July 2012. “I found him fairly impressive although disappointingly (or maybe positively for us) unambitious,” Zuckerberg wrote, explaining that Koum wanted to keep WhatsApp as something minimally updated, like Craigslist. Hansen then asked if Zuckerberg was worried about WhatsApp pivoting to become a Facebook clone. He said no, and that “if anything, once they joined, I basically ended up pushing to add things like” making the WhatsApp status into a stories feed. Zuckerberg was the ultimate decider that pulled the trigger on the WhatsApp deal.

Hansen finished his “cross” by asking about the rate of industry change, to show that tech competition is constantly evolving. Zuckerberg let us know that in the future (the future, Conan?), remote workers will appear in meetings as holograms, and AI bots will also appear as holograms. We’ll be living in an augmented reality world. It wasn’t really clear what the point of these speeches were, and reporters in the press room were commenting that they were repetitive. In visual form, the monologues reminded me of this meaningless Thomas Friedman graph from Thank You for Being Late:

Asking the Burrito Question

Somewhere during the re-direct of Mark Zuckerberg, I couldn’t help but think back to a lesson from early in my career about knowing when to stop with witnesses. So let me spin you this old yarn, and then we’ll look at what came undone today. There’s an old mock trial teaching exercise called the “Burrito Question.” It’s a practice examination of a witness that assumes you have no documents and want to establish what the witness had for lunch. It goes something like this:

Q. Did you have chicken for lunch?

A. I did.

Q. Did it come with cheese?

A. It did.

Q. How about with salsa?

A. It had salsa.

Q. Did it come inside a tortilla?

A. Yes, yes it did.

Q. So you had a burrito for lunch?

A. Actually, it was a taco.

So you can see what went wrong here. It’s a simple illustration of a rule of thumb not to ask at trial a question that you don’t know how the witness will answer, or that you don’t have some material to impeach the witness with. But at a higher level, it’s about knowing when to stop. If you’re trying to prove that the witness had a burrito for lunch, and your standard of proof is, like it is for the FTC here, simply more likely than not, you don’t actually even need to ask the ultimate question. You can stop after establishing that the witness had something inside a tortilla and argue that the factfinder should infer that it was a burrito. Often settling for the inference is better than going for the ultimate admission and getting the opposite answer.

A real-life example comes from my old law firm, Cravath, in defending IBM in a 13-year-long antitrust suit brought by the Department of Justice. In one of the great “coaching trees” in antitrust law, David Boies, the lead lawyer for the government in the DOJ’s case against Microsoft, was the protege of Thomas Barr, the lead lawyer for IBM against the DOJ, who in turn was the mentee of Bruce Bromley, who had cut his teeth on the Famous Players-Lasky case brought by the FTC in the 1920s.¹ Then a Cravath partner, Boies got the call to cross-examine the Justice Department’s lead expert economist, Alan McAdams, at trial. As James Stewart tells it in The Partners, Boies cross-examined McAdams for five days, “destroying almost every point the economist had made during his direct examination.”

Bruce Bromley

But Bromley, in his 80s and nominally the senior relationship partner for IBM, chewed out Barr for Boies’ cross. The government lawyers simply asked questions on re-direct that let McAdams correct the errors Boies exposed. For Bromley, the “more effective approach would have been simply to attack McAdams’ credibility, and use his error in post-trial briefs, when it was too late for McAdams to correct them.” While some of the younger partners dismissed Bromley as a crazy old man, for other IBM lawyers, what he said made sense. Luckily for IBM and Cravath, the Reagan DOJ abandoned the IBM suit on the day it announced the breakup of AT&T.

So how did the FTC clean up Zuckerberg’s cross on re-direct? Matheson tried to a land a complicated point that didn’t really work. He wanted to use Zuckerberg’s testimony about how YouTube has surpassed usage on Facebook and Instagram combined to show that this did not trigger the “network collapse” Zuckerberg feared that Instagram might cause for Facebook. The gist was that Instagram—not YouTube—had the risk of causing network collapse for Facebook because Instagram competed with Facebook and YouTube does not. True to his inner Bill Gates form, Zuckerberg said they had never defined “network collapse,” but that content was shifting to richer forms like video anyway, across platforms.

Then the FTC explored Zuckerberg’s passing on making an offer to buy TikTok—strange if TikTok were actually a competitive threat given the other emails we’d seen. It wasn’t clear what goods the FTC had here since the key part of the document was under seal so not shown in court. Zuckerberg said he didn’t want TikTok at whatever the discussed price was because there was risk from being based in China.

Then Matheson asked about Google acquiring an app, though I didn’t catch what app he was talking about. In any event, Matheson tried confronting Zuckerberg with past testimony here, but that didn’t work either. The FTC seems to be operating under a misunderstanding of a recent amendment to Federal Rule of Evidence 613(b), which tells lawyers how they have to admit a “prior inconsistent statement” from a witness. The new rule says that you can’t admit that statement into evidence “until after the witness is given an opportunity to explain or deny the statement” and an adverse party can question about it. Asked to explain the inconsistency, Zuckerberg dived into a speech about how his two sets of testimony were consistent after all, blowing up the point. This wasn’t a result required by that Rule of Evidence, which doesn’t apply to a statement from an opposing party. So it was a “burrito” outcome—asking the ultimate question in a way that allowed the witness to fix the problem, exactly what made Bromley angry. At one point, Chief Judge Boasberg poured over his rulebook, perhaps investigating this point or the attempts to “refresh” Zuckerberg’s recollection.

The FTC then brought up Zuckerberg’s statement that the number of friends that people add on Facebook was declining, and that Facebook could have done more to address this. Zuckerberg said he could have hired 10,000 engineers and thrown them at the problem, but that it was a cost tradeoff. The answer was similar when Matheson suggested replacing ads with more “people you may know prompts.” Matheson asked a bunch more questions, but now Zuckerberg kept answering that he didn’t remember or didn’t know. More refreshing. The FTC tried to end on a Perry Mason moment, contending that AI “can’t generate content from people in my life.” It was a philosophical point that Zuckerberg answered by explaining the shift in social connection to messaging apps and using a suite of apps to serve a “spectrum of use-cases.” Matheson concluded, and Zuckerberg stepped down.

“I Was Wrong. Very Wrong.”

The FTC’s problems continued after calling former longtime Facebook and then Meta COO Sheryl Sandberg. Taking over for Matheson was the FTC’s chief trial counsel, Susan Musser, who was recently serving as the acting director of the Bureau of Competition. Sandberg is a much different witness than Zuckerberg, willing to answer succinctly and give yes or no answers. She was hamming it up with cutesy asides to Musser and pledges that she was trying her darndest to “be accurate.” Things started off in a promising way:

Q: From the time you started to the time you left, connecting with friends and family was a core use case for Facebook?

A: Yes.

But she became a more difficult witness with documents. In the first email shown, about Google+, she wrote, “[F]or the first time, we have real competition and consumers have real choice.” Musser pointed out that YouTube had by that point been part of Google for a while. Sandberg walked this statement back a couple of times, saying there was always other competition, but that Google+ was a unique threat because it was a “direct, almost exact copy” of Facebook. On cross, she would cite MySpace, Pinterest, Friendster, Twitter, and YouTube, which she added also became increasingly competitive. Why were they competitors? Because there’s a “choice in what you spend your time on.”

Sheryl Sandberg, Credit: World Economic Forum from Cologny, Switzerland, CC BY-SA 2.0 via Wikimedia Commons

The next module made an interesting point: Facebook had a policy over the years where it could refuse to place advertisements for competing firms. We saw an email chain where Sandberg, Zuckerberg, and other executives all agree to block ads from Google+. But that begs the question—if all those other apps were truly Facebook’s competitors, wouldn’t Facebook have invoked that clause to stop ad sales to them? But there’s no evidence that ever happened, supporting the idea that they’re not really competitors. By contrast, the FTC showed a separate email chain where Zuckerberg said he would block WeChat, Kakao, and Line ads, supporting the idea that WhatsApp, as a fellow messaging app, was a similar competitive threat.

Pressed about the reason for acquiring Instagram, Sandberg said the “focus was on the quality of their product, rather than concerns about ours.” She didn’t remember competition from Instagram “being the focus of the discussion.” A missed opportunity here to follow up: even if it wasn’t the focus, do you remember it being one reason discussed?

We had one more burrito problem with the "not exactly killing it" email from yesterday where Zuckerberg says Facebook could scale back Camera, its mobile photo app project, because they were acquiring Instagram. Asked to confirm this, Sandberg says she doesn’t believe the scaling back happened, but agreed Camera was eventually “deprecated.” That was an answer Musser didn’t think she’d get. There was no attempt there to impeach or refresh recollection, which tells me that the FTC had never asked Sandberg that question before. It was a little surprising that the FTC is asking questions they don’t know the answers to this late in the case. Perhaps the FTC was trying to clean up Zuckerberg’s testimony. But it really didn’t need to ask any of those questions. The document was already in evidence and speaks for itself. Musser smartly cut bait on the point and moved on.

The FTC got Sandberg to walk back a few more documents. Was WhatsApp really the “biggest competitive vector” for Meta as Zuckerberg told the board? Answer: “I don’t know that this was the biggest competitive vector; it was a competitive vector.” Ouch. Now we see a line elsewhere in this document from Zuckerberg that he wanted to “use M&A to build a competitive moat around us on mobile and ads.” (Emphasis added.)

At last, Musser brought up the Cambridge Analytica scandal and a presentation to the board of directors on that topic around two weeks after the scandal broke in 2018. The presentation showed that after the news leaked, Facebook’s measure of consumer sentiment—”RCAU”, for relative cares about users—declined, to the “same level as Walmart.” Musser then put up a demonstrative that we’ll see the FTC’s economic expert, Scott Hemphill, use: it compared the RCAU decline to daily active users, which stayed constant. This is a really important point to show that Meta is a monopolist: users have nowhere else to turn to for personal social networking, even when they dislike Meta. If TikTok were truly disciplining Meta, then daily users would drop off and use TikTok instead. But they don’t.

One juicy tidbit, although I’ll bracket here that I might have misheard Zuckerberg. It sounded like he didn’t remember whether he was the final decider on acquiring Instagram. But Sandberg testified that he pulled the trigger. This amplifies the importance of his contemporaneous emails about the rationale—to “neutralize a potential competitor.”

Finally, on cross, in addition to listing out other competitors, Meta’s lawyer asked if Sandberg agreed with her statement to Zuckerberg that $1 billion for Instagram was way too much. “I was wrong, very wrong,” she testified. “If you put Facebook resources behind this product, we could grow it”—that’s what Zuckerberg saw as the product leader.

Cross of Sandberg finished up today (Thursday, 4/17), and then the FTC called Jim Goetz of Sequoia. We’ll bring you Thursday’s summary late tonight or early Friday.

Stray Thoughts

• One interesting comment from Zuckerberg on cross was justifying the acquisition of WhatsApp by the need to have a “suite” of products to stop Apple and Google from bullying Meta on distribution. This was using antitrust offense—Meta’s potential antitrust claims against Google and Apple—as defense. "In retrospect, those platforms have done plenty of things" that have been "pretty detrimental to our business and our community over time,” he testified. I represented the State of Utah in the multistate antitrust suit challenging Google’s Play Store business practices (Utah v. Google), and Zuckerberg would have been a good witness there; he could still be one in the DOJ-led monopolization suit against Apple.

1. Going even further back: Bromley trained with Henry Stimson, who trained with Elihu Root, who trained with John E. Parsons, the lead lawyer for American Sugar in the 1895 antitrust case, United States v. E.C. Knight. Parsons trained at the New York law firm of Benedict & Boardman, founded by a relative of this author’s ancestor.