Summary

Zelnick’s takeover of Take-Two Interactive (TTWO) was “one-of-one” — a hostile takeover executed with no money, because “we had no money.” His fundless firm ZMC found a plain-vanilla unamended Delaware charter, solicited the maximum 10 shareholders allowed without triggering a filing (who held ~70% of stock concentrated among ~20 hedge funds), and won an unheard-of floor vote at the 2007 annual meeting with an 88% provisional count — walking into the corner office the next morning of a company with ~$700M revenue, ~$50M cash, an indicted chairman, and four investigations (IRS, SEC, NY DA, FTC). Market cap then: ~$700M. Today: ~$35B.
The irony that haunts the story: Zelnick built a video game division inside BMG, was forced by new Bertelsmann CEO Thomas Middelhoff to sell it — and the buyer was Take-Two, which a month after BMG sold its 20% stake for $14M released the division’s first completed property: Grand Theft Auto. Middelhoff had told him not to sell the stock in the open market, but Zelnick did anyway. In Zelnick’s view, most people believe GTA is the most valuable entertainment IP ever created — and BMG had briefly owned a fifth of it for a rounding error.
His structural thesis: the movie business has had “heads I win, tails you lose” economics since the 1955 consent decree turned a studio system into a boutique system; video games remain a studio system — “the moral equivalent of the motion picture business in the 1920s.” Games are now the biggest entertainment business — bigger at the software level than all other forms combined and growing at roughly double the rate.
On the AI-kills-gaming thesis that sent the stock down: “hit creation is not asset creation. Asset creation is a necessary but insufficient condition for hit creation.” Data sets are backward-looking, creativity is forward-looking, “clones don’t sell,” and “all hits are by their very nature unexpected” — thousands of mobile games ship yearly, “zero to five” hits emerge, “and you know who makes them? Thank you very much, we do.” Meanwhile Take-Two runs ~200 internal AI projects under CTO David Klein.
GTA 6 is scheduled for November 19, about 18 months behind the original date — and Zelnick backs the talent through slips because that promise “gets tested at the worst possible time.” The template: approving a $50M remake of a finished game two months before release because the creators said it wasn’t good enough — that game became Borderlands. “Culture like character is tested in the breach.”
The operating edge is being boring: “we run a rational organization” — no yelling, no indictments, solid balance sheet, and turnarounds that start by renegotiating the top 10 vendors instead of firing people. He cut ~$40M at Take-Two this way; in a GTA release year the company could make ~$100M and lose money the other three, so the plan was simply “make other hit titles” under the rubric “be the most creative, the most innovative, the most efficient.”
His success meta-lesson is specificity of ambition, not magic: he set out in ‘01 to build a ~$20B tech-supercharged media portfolio with $300K of his own money and offices with water sheeting down the restroom walls — the company, taken as a whole, is now “call it a $40 billion company.” “By knowing what you want and working in service of what you want, you are much more likely to get what you want.”

Deep dive

1. A one-of-one deal: hostile takeover with no money

Zelnick opens with unusual candor about survivorship bias — “stories like this tend to be sort of self-serving because it worked out well… most deals don’t happen, and deals that do happen don’t always work out” — before conceding this one is unrepeatable: “we essentially did a hostile takeover with no money. The reason we did that is we had no money. So it was really our only choice.”
The frame for everything that follows: ZMC, founded in ‘01, existed to buy, turn around, and build companies “at the intersection of media and technology,” on the thesis that technology would supercharge media and both create and destroy value. “I thought in ‘01 it was pretty obvious, but it was not obvious to the entertainment business.”

2. Columbia Pictures, 1983: the least valuable executive becomes the new-media guy

His first job was distributing film to free television — “the last stop on the train” — when “new media” meant videocassettes and pay TV. Columbia’s logic for staffing it: “Who is the least valuable executive at this company that we can put in charge of new media? And that was me… the best thing that ever happened to me.” His analogy: the way today everyone needs an AI guy.
He was a futurist by looking backward — his grad thesis covered electronic entertainment from 1895 (wax cylinders, player pianos) to 1983 — and drew two durable lessons: “you always have to embrace new technologies. If you fight against them, you’ll be left behind,” and “most human beings are wired to believe that which is going on now will never change. But the actual state of play is exactly what is going on now will change.” Politics, the economy, media — “it won’t stay forever.”

3. Vestron to president of 20th Century Fox at 32

After doubling his Columbia division’s revenue in six months, he became president of Vestron — the largest independent home-entertainment company — at 29, three years out of school, and greenlit Dirty Dancing, the highest-grossing independent film of all time until Blair Witch. But he saw the flaw: “You can never have a competitive advantage in the entertainment business through distribution only… you got immediately go into production” — the same reason Netflix produces content.
Joe Roth, an independent producer Zelnick knew from Young Guns, was named Fox chairman in ‘86 and — on Barry Diller’s instruction to “hire a business guy” — reached for the number-one guy he knew, which Zelnick deflates: “I was not the number 100 guy. I was the number one guy in this little narrow independent entertainment business.”
The hiring itself is a period piece: breakfast at the Bel-Air, a same-morning meeting where Diller asked incisive questions and Rupert Murdoch sat silently, then the HR head walking him straight into an office to sign. “It doesn’t happen that way anymore.”

4. What Diller and Murdoch actually taught him

Diller’s method was “robust debate with a lot of heat” — “he’s right 99% of the time. But 1% of the time I’m actually right” — and the survival skill was depersonalizing it: “if you back down, if you didn’t argue, you were done… It wasn’t about me. It was about the thing itself.” A colleague’s image stuck: the inflatable clown doll — “You just get punched, you go over, you keep coming back up.” Zelnick admits he’s collaborative by nature; he comported with Diller’s style to reach right conclusions.
Murdoch’s lesson came during the STAR TV over-leverage crisis — ~150 banks, cross-default mechanics, the whole family business at risk: “He was cool the whole time… laser focused on the conclusion, and he reached the conclusion just by being focused and calm.” Zelnick would later channel exactly this on his own first deal.

5. The blinding flash: he should have run a movie studio in 1927

At Fox he internalized that movie economics have been terrible since the 1955 consent decree converted a studio system (talent on payroll) into a boutique system (talent auctions services per project): “heads I win, tails you lose” — the boutique extracts the upside in success, the studio eats the entire flop. Video games, by contrast, remain a studio system to this day.
He cites a Buffett aphorism he saw in the back of Forbes — a management with a reputation for brilliance meeting a business with a bad reputation, “it is the business’s reputation that will stay intact” — and drew the conclusion: “my ambition should have been to run a movie studio in 1927… What is the moral equivalent of the motion picture business in the 1920s? Well, that is the video game business.”

6. Hollywood’s Atari scar, Rupert’s “no,” and a 95% pay cut

Getting religion on games was contrarian because of the 1982 E.T. debacle: Warner’s Atari built the game in five weeks, manufactured more cartridges than there were consoles, and cartridges were later shown being plowed into a landfill. Warner took a ~$400M write-down. “The rest of the business is like… don’t ever get involved with video games.”
When Zelnick pitched Murdoch on building a games business, Rupert said “Great. Go start one” — but on equity: “No interest. Like, that’s not an offer.” Rupert has no partners except public shareholders. So Zelnick took a 95% pay cut to run Kleiner Perkins-backed startup Crystal Dynamics, moving his family to a rental about to be torn down for a golf course.
His honest self-assessment on entrepreneurial timing: “there are kind of two stages in life to be an entrepreneur. Before you have anything… and after you have some protection. And I was right in the middle… I don’t think I was emotionally ready.”

7. The BMG catastrophe: selling Grand Theft Auto for $14 million

Recruited to turn around BMG — whose revenue was roughly $5B — he negotiated the right to build a games division inside it first: ~$40M, “a rounding error,” using BMG’s worldwide physical music distribution to move cartridges and discs with no added overhead. No other music company was doing it.
New Bertelsmann CEO Thomas Middelhoff — “a person who bet on his gut. It didn’t work out very well for him. At all” — ordered a divestiture before a single title shipped. The buyer, a ~$100M-cap public company, paid 20% of itself in stock; Middelhoff told Zelnick not to sell the stock in the open market, but Zelnick sold it anyway for $14M. “A month later this public company launches the first property that we had completed… The name of the public company was Take-Two Interactive and the first release was Grand Theft Auto.”
Zelnick’s reaction is the tell on his temperament: “I don’t get pissed off easily… I knew it was a foolish decision” — but he left, because you can’t work for a boss who decides that way, and “now I’m ready to be an entrepreneur.”

8. ZMC with $300K, borrowed offices, and a flooded bathroom

ZMC launched with no capital: $300K of his own money, people working free, “one fewer chair than we had people,” restrooms with “water sheeting down the walls.” Capital providers wanted deals first; targets wanted capital first — he realized after watching a PE firm’s pleasantries that “they said it to everyone. They were just looking for free options.”
The lightning-in-a-bottle first deal was Columbia Music Entertainment, a Japanese record company with 0.5% share, too much debt, no hits, “probably criminals lurking around” — a deal he told Ripplewood partner Jeff Hendren to run from twice, until Hendren asked: “Just how busy are you?” It took nine years and was “the hardest deal I’ve ever done,” but channeling Murdoch’s calm — knowing “if someone gave me $300 million to take over a company and I failed, that would be my last deal” — ZMC produced a ~22% IRR in recorded music in the early 2000s. “Unheard of.”

9. Doing free work for Carl Icahn — “you never know where this can go”

Zelnick cultivated Icahn — “the man who said, ‘On Wall Street, if you want a friend, buy a dog’” — by bringing ideas explicitly unpaid. His first, Reader’s Digest, made Icahn about $50M on an approximately $50M position; the reward was dinner at Il Tinello, where a martini glided into Icahn’s hand in one motion and Zelnick’s wife leaned over mid-meal: “Honey, he doesn’t know your name.”
Partner Karl Slatoff, still Take-Two’s president, fumed as Icahn escalated to farming ZMC out for free research to his Wall Street friends — “he has informed us we will never get paid. What are we doing here?” Zelnick’s answer, twice: “You just never know where this can go.” ZMC’s own memos on Take-Two, notably, said stay away — “not being so forward-thinking,” he laughs.

10. “Read the bylaws”: the mechanics of the impossible takeover

Icahn’s cryptic call — “Have you read the bylaws?… You know what you ought to do? Read the bylaws. And he hung up” — revealed a plain-vanilla unamended Delaware charter permitting removal of the board via 50.1% written consent. The distressed stock sat in ~20 hedge funds; SEC rules allowed soliciting up to 10 shareholders without filings, and 10 held ~70%. ZMC couldn’t afford the standard ~$3M proxy fight, so this was the only path.
Commitments topped out at 48% — “I was like, okay, it’s game over” — until partner Ben Feder found a second provision: raise your hand at the annual meeting, win a majority of votes physically present, and take the company on the spot. “Unheard of. It doesn’t exist, but it did exist here.” Then disaster: the hedge funds had lent stock to short sellers and couldn’t vote it. “So what percent do we have?” “You’re not going to like this. 22%.”
At the meeting — journal coverage, TV cameras, the CEO declining to attend — a law-school classmate waved hello, chatted five minutes, and turned out to be Fidelity, a large holder that had refused to join the group and typically backs management. “I looked at my solicitors and I said, ‘We just won.’” The provisional count: “You have 88%. Will you accept that or do you want us to keep counting?” He walked into the corner office the next morning as chairman, in 2007.

11. The turnaround playbook: renegotiate vendors before you fire anyone

What he inherited: one good property, GTA, in development and soon delayed; a tiny profitable NBA franchise; everything else losing money. In a GTA release year, the company could make a meaningful profit of roughly $100M and lose money the other three years. The plan: cut costs, “run a rational company, be friendly to creative talent, diversify the product pipeline, and build an entertainment company.”
His signature move cut ~$40M without initial layoffs: survey the top 10 vendors, negotiate them all down — you save money immediately, you don’t scare the team, and “the team’s like, wow, they just saved all this money and no one got fired.” Never open with expense-report policing: “that’s a great way to alienate David. And if David has another option, he’s going to leave the next day.”
A ZMC rule worth stealing: only do turnarounds as the first team in. “We’re not so smart. Smart people all do the same things” — a prior failed turnaround probably already tried everything you would. And on early Take-Two mistakes: “How could you? When you’re on the floor, there’s nowhere to fall.”

12. The pitch to talent — and the $50M Borderlands test

The must-have pitch wasn’t rationality; it was “we are going to give you the resources, creative and financial, to pursue your passion… and support you through thick and thin” — then, as the nice-to-have, a grown-up company: no screaming, no indictments, no getting sideways with the FTC, and eventually a balance sheet so “when we fail — inevitably, you fail in the entertainment business — we can live to play another day.”
The test came pre-turnaround with limited capital: a division head said a game two months from release wasn’t good enough and wanted a remake — $50M incremental and another year. Zelnick dug in, then supported it. “That title became Borderlands… I’m pretty much going to assure you, no one else in the business would have done it. Why? Cuz it was insane.” His maxim: “Culture like character is tested in the breach.”
On managing disagreeable creative geniuses: “I love them… you have to sincerely care about other people to be an effective leader” — but not family (“my family’s at home. This is an enterprise”), and with limits: one abusive creative’s emails he simply stopped reading — “the delete key is there for a reason” — until “eventually he had to go.”

13. GTA 6, November 19 — and why games are the biggest entertainment business

On the endless “when’s GTA 6 coming out”: “November 19th, I do know” — about 18 months behind the original date. On whether GTA is the biggest game ever: it depends how you count against Mario Kart and Call of Duty franchises, “but in terms of the value of the IP, most people believe it’s the most valuable entertainment IP ever created.” GTA 5’s total revenue is undisclosed: “It’s a lot.”
He describes Take-Two as a media and entertainment company: “We provide an entertainment experience expressed through video games.” According to Activate Consulting, the U.S. media day is around 13 hours; anything within it can compete or coexist, though playing a game remains different from watching a linear show. At the software level, games are bigger than all other forms combined and growing at roughly double the rate. His 90-year-old mother plays bridge online with friends she’s never met. He’d consider diversifying into growth areas like live entertainment, but “we would not diversify backwards into backward-looking legacy declining businesses.”

14. Visualization isn’t magic — it’s forced specificity of ambition

Zelnick’s distillation from “a billion biographies”: the single common factor among the highly successful is “they knew what they wanted.” Not Napoleon Hill mirror-writing — “by knowing what you want and working in service of what you want, you are much more likely to get what you want.” Magical thinking, by contrast, is “because I want it, it’s going to happen.” Senra concurs from his 415-biography database: if he had to compress a decade of study into one word, “it’d be focus.”
The arc: he set out in ‘01 to build a ~$20B company; at MIT ten years in, asked if he was “deeply disappointed” at ~$1B, he answered “it’s directionally correct and I’m not stopping.” The company, taken as a whole, is now “call it a $40 billion company. We kind of exceeded our goals and we’re not done yet.” Take-Two’s creed — mission: number one entertainment company on Earth; strategy: most creative, most innovative, most efficient; culture: excellence, teamwork, kindness — is knowable by all 14,000 employees.
His mentoring surprise — he spends 20–25% of his time coaching and says he is easy to reach, though people still need to do homework: about half the people he assigns exercises never follow up. “There are people who believe if I can just get into a room with that really successful person, they can just wave a magic wand.” His advice to a new employee: show up Monday asking how you’ll create more value than you cost, and audit yourself Friday.

15. Service beats self-promotion — but one size doesn’t fit all

How to Win Friends and Influence People “changed my life” at Fox, where he “nearly failed” running the subtext “I’m the smartest guy in the room… a couple of Harvard degrees, we’re just going to do it my way” — masked behind humility he couldn’t fake, leaving him anxious. The shift: “How can I be of service to you?” He retells Carnegie’s postal-clerk story approvingly: “Are we so small-minded that we need to get something out of every interaction?”
But he refuses to universalize it: “Elon Musk’s approach to business is diametrically different than mine… last time I checked, he’s the richest guy on Earth and I’m not” (per the new Musk book, “camaraderie’s dangerous”); Diller’s Socratic method “can be brutal” and works for him. His own confessed cost: “I can sometimes support people who are on the team longer than I should… there are times when I’ve kicked the can down the road.” Senra’s addition, via Daniel Ek: young founders waste years playing Jobs or Musk instead of finding their own archetype.

16. AI creates assets, not hits

Take-Two runs enterprise versions of ChatGPT and Claude plus ~200 productivity projects, all within technology under CTO David Klein — no separate “AI czar,” because “we’re a technology company that makes entertainment. We always have been.” Internal criticism, notably, is that they’re not moving fast enough.
His rebuttal to the claim that sent the stock down (“with AI, anyone can make a video game”): “Anyone can make a video game last week… Thousands of mobile games get put out a year. You know how many hits are made in a year? Zero to five. You know who makes them? Thank you very much, we do.” AI is “big data sets, lots of compute, and a large language model mushed together” — data sets are backward-looking, creativity is forward-looking.
The load-bearing distinction, kept verbatim: “Hit creation is not asset creation. Asset creation is a necessary but insufficient condition for hit creation… You can clone GTA in 3 years with pre-AI technology. But it won’t be GTA. Clones don’t sell. All hits are by their very nature unexpected.” He hedges the upside honestly: who would benefit more from AI-made hits than the company that already owns the IP — “but that doesn’t mean AI isn’t super helpful.”

Freda's Investment Notes Ep. 2: Tokenmaxxing and Human Connection Anthropic Partners With SpaceX AI, Leopold's $5.5B Bet, and the Singularity Economy | EP #255