David Senra · JA + tiny EN

ブートストラップでFCF50億ドルへ——AppLovinのアダム・フォロウギ

Bootstrapping a Business to $5 Billion in Free Cash Flow | AppLovin’s Adam Foroughi

英語はYouTube自動字幕ベース。日本語は学習用の整文対話。一次情報は動画。

センラフォロウギ
FAQ 一覧 YouTube
センラ#1

何十億ドルもの自社株買いをやった話を聞かせてくれ。

EN

Tell me about these billions of dollars of stock buybacks you did.

フォロウギ#2

上場したときの文脈から話す。2021年4月、COVIDの真っ只中で上場した。時価総額はだいたい280億ドル。最初の半年で400億まで上がって、みんな興奮した。そこから2022年は文字どおり毎日下がった。底は時価総額38億ドルくらい。株価は115ドルから9ドルまで落ちた。事業を回しているのに、世界中が「その事業はゴミだ」と言っている状態だ。

一方で、2021年のEBITDAは7億ドル。2022年は10億ドル超。成長は速く、事業はちゃんと実行している。なのに公開市場の投資家は「ひどいビジネス」と言っている。

EN

Uh so let me give you a little bit of context on when we went public and why this became a pretty big opportunity. We went public in COVID in April 2021. Company was worth about $28 billion. Went up to 40 billion first 6 months. We're all excited. Stock goes down from there literally every day for all of '22. We got to a floor of I think like $3.8 billion. We went from $115 a share to $9 a share. So you're running a business and the whole world is telling you your business is trash. Like Like what What do you do? At the same time, interestingly, we went out with $700 million of EBITDA in '21. We did a billion dollars plus of EBITDA in '22. So like fast growth, executing on the business. Yet public market investors were telling us business is terrible like and and it was what it was.

センラ#3

なぜそう見られた?

EN

Why did they think the business was terrible?

フォロウギ#4

わかりにくい領域だからだ。広告とゲーム——公開市場で難しい二つの場所にいる。もう一つはCOVID期にIPOが多すぎたこと。公開市場に4年いて学んだが、初期には大口投資家が株を買って支えてくれないと厳しい。プライベート時代の投資家は売りたい。COVIDのIPOラッシュでFidelityやBlackRockのような大口は、新IPOを一つひとつ精査できなかった。見分けがつかないほど出た。

結果、支えのないキャップテーブルになった。株が市場に流れて崩れ、毎日下がり、「この会社、何かおかしい」と見える。株価を越えてファンダメンタルズを見ようとするのは難しい。落ちるナイフを掴むようなものだ。92%下がって「終わりにしよう」と反応・防御に入るのは簡単だ。

僕らがやったのは二つだ。世界が株を嫌っているなら、自分たちで買い始めればいい。底で時価総額38億、EBITDAは10億超。キャッシュフロー対時価はだいたい5分の1。本気なら理論上、1年で発行済みの20%を買い戻せる。そこで買い戻しを始めた。

やり方は普通と違った。多くの会社は公開市場で誰から買うかわからないまま買い戻す。僕らはキャップテーブルの約50%が、数年のうちに売るとわかっていた——PEと、もういない創業者たちだ。だから一般市場ではなく、売る側と合意して時間をかけて買い戻した。その後18ヶ月で、自社資金とレバレッジを合わせてだいたい60億ドル規模の買い戻しを実行した。結果としてその買い戻しは、だいたい500〜600億ドル相当の価値創出になった。会社史上でもかなり成功した買い戻しのひとつだ。

EN

It is a tough space to understand. First of all, like we were We're in advertising and we're in gaming and those are two tough places to be in the public markets. The other challenge was that we went out in COVID and there were just way too many IPOs. And once you learn the public markets, which I've learned a lot more um as we've been public for now 4 years, you need really big investors to start buying shares in companies early on cuz you have your private company investors trying to sell shares. So on the one hand, you have this imbalance. You've got a flood of shares that are going to come to the market. And with all the COVID IPOs, the big funds, the companies like Fidelity, BlackRock, etc., they weren't doing as much research on any new IPO cuz they couldn't tell the difference. There was just too much hitting the market. So we ended up with a cap table that didn't have support. And then shares started selling into the market, stock crumbles. You start going down every day, people look at the company and go, "This company's Something's wrong with this company. Like what's happening?" And it's really hard to look past the stock price and go, "Let me look at the fundamentals and try to assess what's going on." And and yeah, you're you're out there. It's like catching a falling knife. Like what do you do as an investor? So we were thrown out. It's easy to go, like, we're down 92% hang it up, like, what are we going to do? Start getting reactive, defensive. Um, instead of that, what what we ended up doing was going two things. One is the whole world doesn't like our shares. So, if no one's going to buy our shares, why don't we just start buying our own shares? And so, we kicked off a really successful buyback. The company at the bottom was worth $3.8 billion market cap and we were generating over a billion dollars of EBITDA. So, you think about the cash flow to to market cap, it was about 1/5 that we were generating every single year. Well, in theory, we could buy back 20% of the shares of the company just in the next year if we really believed in the path we were on. So, we kicked that off. But, we did it a little bit differently than what most companies do. Most companies go out and say, "I'm going to buy shares from the public markets and just take shares back." But, you don't know who's on the other side of that trade. On the other hand, we knew that we had a cap table where about 50% of the shares were going to sell at some point over the coming years. We had private equity investors that owned roughly 50% of the shares of the company alongside some other founders that were no longer there. So, instead of going to the public, we went to the shareholders that we knew were going to sell and got them to agree to sell back to us over time. And so, for the following 18 months, we ended up deploying somewhere around $6 billion of buybacks of our own capital and we levered some to buy back shares in the company. And over time, that ended up creating somewhere in the neighborhood of 50-60 billion dollars of actual proceeds from the buyback, one of the more successful buybacks in the history of the company.

センラ#5

約60億買って、その60億で約600億。一部は利益、一部は借り入れ?

EN

You bought back around 6 billion, you made about 60 billion on that 6 billion. Where did you get that you said some of it came from earnings of the company and then you you went and

フォロウギ#6

レバレッジも使った。

EN

We levered as well.

センラ#7

どこから借りた? 当時、時価総額38億で借りて買い戻すなんて狂ってる、と周りに言われたはずだ。社内の見え方と、世界が事業について言っていることの差を説明してほしい。

EN

Where did you borrow the money from?

フォロウギ#8

現金は保守的に持ちたがる人が多い。レバレッジ自体が怖いのに、「会社はクソだ」と言われている株を借りて買うのは本当に怖い。僕は「雨の日のために現金を貯める」派じゃない。僕らが作っているものと行く先を信じている。高キャッシュ創出なら、常に自社株を買い戻すべきだ。バリュエーションが甘い底で恐れる理由はない。

取締役会の経験は当時あまりなかったが、ボードは支持してくれた。キャッシュフローの5倍のマルチプルを見て「買えるだけ買え」——ロケット科学じゃない。安くて成長に確信があるから、できる限り叩いた。

EN

So, as a private company

センラ#9

ボード経験が少ない、とは?

EN

tell you at the time, "You're nuts for borrowing here. You're down 3.8 billion market cap, right?" Explain like how you were thinking the difference between how you were thinking about inside the company versus, you know, the world telling you about your business because there had to be a lot of people around you saying, "What the hell's wrong with you? You got to be nuts going into debt to buy back your stocks in this shitty company."

フォロウギ#10

創業時、VCに資金調達を話しに行った。トップVCは全部断られた。2012年春ごろ、バリュエーション400万ドルに対して100万ドル——会社の4分の1を渡す話だ。

EN

I mean like a lot of times like people like to be conservative when it comes to cash. And so one to to lever up is scary to people, but then two to lever up to buy your own shares when everyone's telling you that your your company is a piece of … that's really scary to do. I never believed in saving cash for a rainy day. I feel like I'm a big believer in what we're building. I believe in where we're going. So, if I believe in the future and we're really high cash generative business, we should always be buying back our shares. So, at a bottom point where the valuation became that juicy, there's no reason to be afraid of it. And the good news is I mean I don't have a lot of experience with boards and we can talk about that in a bit, but our board was very supportive because it's not it's rocket science. You look at a a multiple of five times cash flow and you go, "Okay, why don't we just buy all the shares? Buy as much as you possibly can." And so, because it was so cheap and we had a lot of conviction on our future growth prospects, we just hit it and hit it as hard as we could.

センラ#11

それが最初のラウンド?

EN

Explain what you mean about you didn't have a lot of experience with boards.

フォロウギ#12

全部ノー。広告とゲームでピッチが難しかった。これが僕の3社目だ。

EN

When we first started the business, I went out and talked to VCs to raise capital and I got turned down by all the top VCs. So, a million dollars over a four million dollar valuation at some point in spring 2012.

センラ#13

2012年? 今なら広告・ゲームは世界で最も儲かる事業のひとつだが、当時何が起きていなかった?

EN

Wait, that was your first

フォロウギ#14

VCにとっての難しさは、「GoogleとFacebookとAmazonがいるのに、変な名前の小さな会社が巨人と戦えるのか」だった。論理の失敗は、僕が広告で3社目で、すでに10年以上前から成功していた点を見ないことだ。人に賭けろ。100万対400万で25%はかなり良いディールだ。モバイルゲームとモバイルアプリの市場自体も、当時は今ほど大きくなかった。僕らはその市場に賭けた。いま巨大なモバイルゲームは当時は小さかった。VC側の見落としだ。

EN

All rejected. That's what I pitched. I would have given up a quarter of the company for um a million bucks. What would have would have been a big payday for those VCs.

センラ#15

前の広告事業はデスクトップ?

EN

And they said no because again we're in advertising and we're in games and it was a tough pitch and um this is my third business.

フォロウギ#16

そうだ。

EN

What what year is this though because I think people 2012 cuz people hearing you say we're in advertising games, now those are some of the most profitable businesses in the world. What wasn't happening in 2012 though? What was

センラ#17

その経験が、デスクトップからモバイルへのトラフィック移行を最初から見えた利点になった。

EN

challenging for VCs back then was well, Google and Facebook and Amazon are in the space. And so, why would some goofy name little company that's trying to develop something be able to compete with the giants? I think the the sort of failure in the logic is cuz it's my third business in advertising. I already had successes over a decade before in advertising. So, it's a failure on seeing who's the the founder. So, bet on the person. A million over 4 million, you get 25% of the company. It's a pretty good deal. And then understanding that the space itself in in mobile gaming and mobile apps at the time was nascent. It hadn't become as big as it is today by any stretch of the imagination. So, you had this like We Well, where we got sort of lucky is we bet on that space, but this gaming space on mobile that became massive now was very little back then. And so, those were misses on the VC side.

フォロウギ#18

運もあった。2005年にソーシャル上の広告事業、2008年にもう一つ。デスクトップのソーシャルにいて、モバイルのApp Storeが立ち上がり、トラフィックが急速に移るのに誰もまだ語っていない——市場についての大きなデータ優位だ。2010年、前の2社を売却してPalo Altoに行き、「モバイルは大きくなる。何をやるかはまだわからないが、モバイルは大きい」と言った。

経済的には成功していたので、広告はもうやりたくなかった。広告はキツい。

EN

Your previous advertising business was desktop-based.

センラ#19

なぜ?

EN

Yeah.

フォロウギ#20

20年広告の人でいることに興奮しにくい。いまは人生コミットで、うまくやれているが、当時は「広告は勝った。DTCでもやってみよう」と思った。

EN

So, the advantage that you had here starting this other company, which is Apple 11, right? Is you saw the shift uh from traffic from desktop to mobile at the very beginning.

センラ#21

B2BからDTCへ行きたかった理由は?

EN

Yeah, we we so we sort of got lucky in that, too. We had uh built a a business on social media in 2005 in advertising with a few other guys. And then in 2008 had another one. And so, when you're on desktop and you're in social and then the mobile app store launches and you start seeing your traffic shifting really quickly, but nobody's talking about it. It It's a huge data advantage on a market. And so, what ended up happening was in 2010, I had sold those two prior businesses with others and just went over to Palo Alto and said, "Mobile's going to be big. I don't know what we're going to do mobile, but mobile's going to be big." Funny enough, because we had had quite a bit of economic success, I didn't want to do advertising again. Advertising is a tough deal. And so,

フォロウギ#22

隣の芝は青い。広告は征服した気になっていた。現実は、当時想像したよりはるかに大きな広告事業を作れると証明した。でもB2B広告では「クライアントは常に正しい」と訓練される。消費者を持てば運命を自分で握れる。だから消費者事業の方が楽しいと思った。

2011年、出会い系アプリとファッションアプリを出した。ひどかった。

EN

Why?

センラ#23

既婚者が出会い系を作り、会うたび同じ服——ファッションセンスゼロの人がファッションアプリ。

EN

It's just really hard to get excited about being an advertising person for 20 years. So, now Now, I'm sort of life commitment to it and I'm pretty good at it, but at the time I was like, "Look, we succeeded in advertising. Go try direct-to-consumer something. And so, why why wouldn't we?

フォロウギ#24

服はだいたい4パターン。当時既婚だった。ひどいアイデア二つ。Tinder前の出会い系、ファッションはPinterestが立ち上がるころで、彼らは近所にいた。市場に機会はあったが、僕らが正しいチームじゃなかった。

EN

And why would you want to go from B2B to direct to consumer?

センラ#25

モバイルに全部が移るのを見て、とにかくアプリが欲しかった?

EN

The grass is always greener and everything. And so, it felt to me like I'd already conquered advertising. What why do I need to do it again? And that's so far off reality cuz I've obviously proven you can build much bigger advertising business than I imagined at the time. But I felt like it's tough. You're dealing with the client on the other side. And when you're in the advertising business or any B2B business, you you're trained clients always right. So, you're on the other side of that trade. If you own the consumer, well, then you own your destiny. And so, I felt like it would be a lot more fun to build a consumer business. And we tried. We we built in 2011 we launched a dating app and a fashion app. Look terrible.

フォロウギ#26

オーディエンスを先に持ちたかった。3つ目に出したのが、初期のAppLovin——アプリ発見アプリだ。2011年晩夏〜秋。つながって「AdamがWords With Friendsをやっている。一緒にやろう」と教えるだけ。アプリ自体はダメだったが、そのプッシュの反応率は天井抜けだった。乗っている人は次々ダウンロードする。

「App Storeは思ったより大きくなる。コンテンツは山になる。アプリはダメだが推薦アルゴはすごい」。それが今の本体になった。推薦アルゴを他アプリ上の広告プラットフォームとして出した。

---

EN

So, a married guy is building a dating app. And you every time I see you, you're in the same clothes.

センラ#27

前の2社を売ってDTCを試し、AppLovinの前身アプリを作った。当時もうお金持ちだった?

EN

Yeah, you made a joke about that earlier. I have like four outfits. I have zero fashion sense. And I was married at the time. For two terrible ideas. Pre-Tinder though. And the fashion app we launched right around when Pinterest launched. And they were down the block from us. So, there was obviously opportunity in those markets. We were not the right teams to do it.

フォロウギ#28

もう1ドルいらなくてもいい、と思えるくらいには豊かだった。残念ながらそのラインは豊かになるほど上がるが、当時は「もう1ドル稼がなくても大丈夫」だった。最初の1年半はブートストラップで試行錯誤した。

EN

Well, you just wanted to have an app because you saw everything shifting there?

センラ#29

ブートストラップは望んだから? それともVCが取らなかったから?

EN

Yeah, I wanted to have an app because I wanted to own the audience before we got into anything else. And so, actually like that that transitionally that goes to the third app we launched was the first version of It was an app discovery app. And this was I think it was in in late summer to fall 2011. You and I would go connect on this app AppLovin. And it would tell you, "Hey, Adam's playing Words With Friends. You should go play Words With Friends with him." And that was the entirety of the app. It was just an app recommendation app. The app itself stunk, but when you got that push, the response rate was through the roof. Everyone who was on that app was going and downloading other things. So, we're like, "Okay, well, there's a potentially this app store is going to be bigger than people realize. There's going to be a ton of content. This app stinks, but this recommendation algo is really cool. And so that that's what really turned into what we became, which is we took that recommendation algo and just launched it eventually as an advertising platform on other apps.

フォロウギ#30

それはもっと前の話。自信のないアイデアを投資家に売りたくなかった。チームとフィットを探すあいだ、ピボットだらけの時期に投資家とやりたくない。初期はLLCにして、100万ドル入れて負けても税効果で50万戻るような形で、他の所得に対してブートストラップした。プロダクトマーケットフィットまで1年半、エンジニアのコストをそれで賄った。広告事業になるとわかってからVCに行った。

EN

One of the things I love most about the conversations I have with Adam is how focused he is on making the best possible product he can for his customers. "Obsess over customers" is a maxim that is repeated by Jeff Bezos, and you definitely see that with Adam. It is also the same trait that I see in my friend Kareem, who's the co-founder and CTO of Ramp. Ramp is the presenting sponsor of this podcast, and Kareem is one of the greatest technical minds working in finance today. Kareem is obsessed with crafting a high-quality product and using the latest technology to constantly create better experiences for his customers. Ramp has one of the most talented technical teams in finance, and they use rapid, relentless iteration to make their product better every day. Ramp is completely committed to using AI to make a better experience for their customers and automate as much of your business's finances as possible. Today, Ramp helps your business control and automate spending. They do this with corporate cards for every single one of your employees, expense management, bill pay, procurement, travel, budget controls, accounting automation, and AI tools that flag waste, fraud, duplicate charges, and policy violations. Tomorrow, Ramp is full self-driving money for your business. Many of the fastest-growing and most innovative companies in the world are running their business on Ramp. I run my business on Ramp, and you should, too. Make sure you go to ramp.com to learn how they can help your business save time, save money, and grow revenue. Let AI chase your receipts and close your books so you can use your time and energy building great things for your customers. Get started today by going to ramp.com.

センラ#31

前の2社もブートストラップ?

EN

You said you had two other advertising businesses. You sold them. So now you're going to try to go uh direct-to-consumer. You're going to build these apps the precursor to to AppLovin.

フォロウギ#32

1社目は投資家あり。3社とも共同創業者がいた。1社目はジュニア共同創業者。2社目は共同創業者で、ソーシャル広告を出してすぐ黒字だったのでブートストラップ。

EN

Were you already rich at the time?

センラ#33

アプリ推薦のアイデアのあと、次は?

EN

I was rich enough to not need more money in my mind. Um unfortunately money that the line on that always goes higher as you get richer, but I felt like if I didn't make another dollar I'd be fine. So I was bootstrapping this business for the first year and a half while we were just tinkering with things.

フォロウギ#34

アプリ自体は誰も使わない——ショックだが、人が使うのは人が写っているアプリで、これは「アプリのアプリ」だった。でも推薦で高確率ダウンロードするなら、自前アプリのディストリビューションは要らない。出てくる他アプリの中に技術を置けばいい。それで広告プラットフォームを作り、SDK——アプリに入れて広告にアクセスするコード——を雇って組み、2012年3月にローンチした。発見アプリは捨て、AppLovinはアドネットワークになった。

円環だ。負けが嫌いなのにDTCでは何もわからなかった。出会い系もファッションも本当にダメで、プロダクトがダメだと負けはキツい。一方で推薦の強さと市場の大きさを見て、広告から逃げる理由がなくなった。「やろう」。パッケージして市場に出した瞬間から、事業はすごく速く伸びた。

EN

Were you bootstrapping because you wanted to bootstrap or bootstrapping because the VCs would not take

センラ#35

ディストリビューションはどう取った?

EN

No, it's pre pre pre pre dated that. I didn't believe in selling to investors an idea that I wasn't confident. So I figured if I'm out there with this team and we're just playing around and trying to find a fit. During that point in time where I'm going to pivot a bunch, I don't want to deal with investors. So we actually set it up as an LLC um early on and the nice thing is you put a million dollars into an LLC, you lose a million dollars, you get back 500 grand. So we had we had a tax efficient way to bootstrap against whatever earnings we had elsewhere. So we did it and we did that for we covered the cost of the engineers at the time did that for the year and a half until we landed on product market fit. I didn't go to VCs until I knew we're going to be an advertising business.

フォロウギ#36

他のアプリ開発者が金に飢えていた。2012年、App Storeはカレンダー、おならボタン、懐中電灯、ゲームの始まり——Candy Crush、Clash of Clans、もっと単純なゲーム。開発者は金が要る。「カレンダーで100万人いるなら、1日1万ドル払う。広告プラットフォームを入れて」と行った。広げ方は僕らの責任だ。

EN

Were the other two companies you started before this? Did you also bootstrap those or you raise money for that too?

センラ#37

彼らの代替は? あなた以外の選択は?

EN

One had had uh investors. And these were both I've had co-founders at all three businesses, but one of the first one had investors and I was the junior co-founder. The second one I was a co-founder at. Um we had just bootstrapped cuz we launched an advertising business on social media and it was profitable right away.

フォロウギ#38

Googleが2008年にAdMobをほぼ10億ドルで買っていた。当時としては巨大ディール。代替はGoogleの広告か、この新しい広告か。2008年買収から2012年まで、そこはあまり革新が進んでいなかった。市場に出たとき、その時点ではGoogleに勝ちやすかった。

EN

Okay. So now you have this idea, you have this app recommendation app. What happens next?

センラ#39

どうやって?

EN

So nobody's using the app which shocking people use apps with hot people in them and this was an app of apps. But we realized if you're going to download at a really high rate based on a recommendation we can send you, well what we don't need distribution on the app. Why don't we just go distribute our technology inside all these other apps that are coming into existence. So what we not rocket science back then, we just realized let's go build an advertising platform around this construct. And so we did that, we we hired a team to build what's called the software development kit, but just code that someone can put in their app to access an advertising platform. And we launched that in March of 2012. So, scrap the app discovery app. AppLovin became this ad network. We launched this ad network. We're right back to ads. But like it what was interesting, I mean full circle one, I didn't like to lose and when we did direct to consumer, it's like I've no idea what I'm doing. The the the not in the dating, not in the fashion, all true. We had no idea how to take to to market a direct to consumer business. And we didn't know what we were looking for. The product stunk. So, it was really difficult to lose. And then second, when we saw how powerful these recommendations were and knew that this market was going to be big, there's no reason to shy away from ads. At that point we're just like let's go. And we packaged up the tech, put it live, and then from that moment that we got into market, business started growing really quick.

フォロウギ#40

最初に市場に出る製品はブランド向けに作られがちで、広告の大金はブランドにある。みんな「ブランドから金を取り、スプレッドを作り、広告を置く」と頭に入れる。当時のAdMobはそうだった(今は違う)。僕らはブランドに売りたくなかった。弱い一考はあったが、売りにくい。初期にニューヨークへ行って代理店と話すと地獄だ。予算の一部にアクセスしてもらい、効果が科学的でなく手振りになる。一番大きい取り分はクライアントを接待する側に行く。その事業にいたくない。営業なし・高価値プロダクトの事業がいい。

AdMobとの違いは開発者自身に行ったことだ。「アプリ内に広告を置かせて。一方でカレンダーやゲームやおならボタンにユーザーが要るなら、うちのプラットフォームで広告を回してユーザーを取る」。最初から今まで、顧客がパフォーマンスベースで買えて、広告費がよく使われ、反対側でマネタイズしてスプレッドを作り、プラットフォームをアービトラージする会社になればスケールできる。ブランド向けではなく開発者向けツールを作った。それが効いた。

EN

How did you find distribution for it?

センラ#41

当時、同じことをする他のプラットフォームはなかった?

EN

These other app developers were desperate for money. So, at the time I mean it's like 2012. The App Store was full of calendar apps, fart button, flashlights, and the beginning of games. You had Candy Crush, Clash of Clans, and some simpler games at the time. Not many though. And these developers needed money. And so, we just went out and let said, "Look, you've got a million users on this calendar app. We'll pay you 10 grand a day. Just put our ad platform in there." It was up to us to make it spread on it.

フォロウギ#42

同時に数社が生まれていた。独立に同じアイデアにたどり着いた。面白いのは、僕を断った同じVCが、それらの会社に大金を入れたことだ。動機になった。あいつらを潰すのが目標だった。

EN

What was their alternative? Like what if they don't go with you, what else could they What other decision could they make at the time?

センラ#43

前に2社成功しているのに断られ、成功歴のない競合には金がつくのはなぜ?

EN

Google had bought this company AdMob in 2008 for I think it was nearly a billion dollars. Huge transaction back then. And so, that was the alternative. Go with Google's ad platform or go with this new ad platform. Since they'd bought it in 2008, when we started in 2012, there wasn't a lot of innovation there. So, when we went to market, it was easy for us to go out compete Google at that point in time.

フォロウギ#44

営業が下手なのかもしれない。あるいは市場がもう少し進んだタイミングだった。VCは市場に強い確信が要る。2012年初は形成中で、僕らは製品を出したばかり。他社が2012後半〜2013に調達するころには、急成長がかなり見えていた。2012年3月ローンチから、11月のエンジェルラウンドまでに収益が立ち上がった。

EN

How?

センラ#45

エンジェルで何をした?「それが全部」とは?

EN

They first come to market with a product that was built for brands and most advertising businesses are built for brands and that's where the big money is in advertising. So, people get get in their mind that we're just going to go take money from brands, create a spread, place ads, off you go. That's what the AdMob business was back then. It's a lot different today. When we went into market, we didn't want to sell to brands. So, it was never a consideration for me. I mean, it we can talk about it, but it was like the one weak consideration. It is a tough thing to sell. I I went over to to New York early days. Everyone told us we should we should worry about brands, too. And try to talk to some agencies and it's brutal. You're trying to convince someone to give you access to part of a budget and then proving that their dollars were well spent is completely hand-wavy. There's nothing scientific about it. It's you're placing their ads, you're trying to get a cut of their budget, and the biggest cut goes to the folks that are wining and dining the client the most. And I don't want to be in that business. I wanted to be in the business of no sales, high-value product. And so, what was different about us to AdMob was we went to the developers themselves. We said, "Look, on the one hand, like allow us to place our ads inside your app. On the other hand, you need more users for your calendar app or for your game or for your fart button or whatever it was, run an ad on our platform and we'll get you users on the other side." So, the very very beginning was and and this has always held true. If our customers are able to buy on a performance basis and we can define what that means in a little bit, but if they're able to buy on a performance basis, their business is getting better because the ads are well spent. And if you spend your money well on ads and on the other side you can monetize, you create a spread, they become effectively a company that can arbitrage our platform, then they could scale their business. And so, we built tools for the developer instead of building tools for the brand and that really served us well at that point in time.

フォロウギ#46

2012年11月、pre 2500万ドル相当のコンバーチブルノートで400万ドルを調達し、自分の金もマークした。その時点で事業は月100万ドルのランレート。ローンチから8ヶ月で月100万・黒字。

EN

That's a really interesting insight. And so, there was no other alternative for the game developer at the time or the app developer rather, that there was no other platform like yours doing what you were doing.

センラ#47

その100万のうち利益は?

EN

There were a couple other companies that were just starting to exist, too. It was interesting cuz

フォロウギ#48

わずか。少し黒字。今のARRランプやバリュエーションは別世界だが、当時モバイルでゼロから年換算1200万まで短期間は本当にすごかった。だから11月の投資家集めは簡単だった。長く知っている人や役に立ちそうな人、自分のドルをマークするため。そのラウンド以降、本格的な追加調達はほぼしなかった。流動性のラウンドは後である。

ボードの話に戻ると、2018年までボードがなかった。

EN

They had your same idea, though.

センラ#49

6年後?

EN

At the exact same time, a few of us had the exact same idea. So, you had this

フォロウギ#50

そう。KKRのラウンドをクローズしたとき。長所は、株主証明書の全部に自分のサイン——社長・副社長・全部——で、全決定が自分を通る。短所は、資本市場でミスをした。ボードと経験者がいれば避けられたはずだ。

EN

Independent of one another.

センラ#51

例は?

EN

Totally independent. And it's also funny, the same VCs that rejected me put a lot of money into those companies. And that was very motivational. We had a as a goal to put those companies out of business.

フォロウギ#52

2014〜15年、すごく速く伸びた。ローンチから3〜3.5年でEBITDAだいたい5000万ドル規模。僕はもう十分豊かだったがチームはまだ。人は少なく急成長で、他のテック企業から買収の関心が出た。現金6億ドルのオファーまで行った。相手名は言わない。自分で事業を回す理由のひとつは、相手側の社員として働く想像がつかないことだ。

当時は全部が僕を通るハンズオン経営で、前年比100%成長も続いていた。タームシートで「6億、現金」と出ると、チームにとっては人生が変わる数字だ。でも数字を見てからクローズまでのあいだに事業がまた伸びる。ファイナンス畑として、やる数字はあったが、その数字ではなかった。だから断った。

EN

Why do you think the VCs turned you down when you had two successful advertising businesses before, and yet they're funding the competitor of yours that has the same exact idea without the the the previous success in there?

センラ#53

その数字は?

EN

I mean, look, I'm terrible at sales. So, maybe it's that. Um, but maybe it was that this market was a little bit further along at that point in time. Like, I think VCs tend to look at a market and they have to really have high conviction in it. And you had in early 2012, things were just starting to form, and we were just getting to market with our product. By the time these other companies were raising in later 2012 and 2013, it was pretty clear this thing was going to grow pretty quickly. And then our our revenue ramp at the point in time when we launched in 2012, from March to to November when we did our we did an angel round, and we marked our own money in, and that was it.

フォロウギ#54

かなり高く、その時点でだいたい10億ドルは要った。

EN

What do you mean that was it?

センラ#55

差額はまだ4億? 今の時価総額は?

EN

So, we raised four over $25 million pre as a convertible note to common in in November. The business by that point in time had gone to a million-dollar-a-month run rate. So, we went from launch to a million-dollar-a-month run rate and profitable in in a matter of eight months.

フォロウギ#56

いまだいたい1400億。そのミスをしていたら、かなりお金をテーブルに残していた。多くの事業は売ると革新の機会を失う。そこで売っていたら、何になるかは誰も知らなかった。確信があってそのレンジを取らなかったのは、ある意味運がよかった。

---

EN

How much profit do you think you were making on a on a monthly basis out of that million?

センラ#57

資本市場のミスに戻る前に。Raphが君についてのノートをたくさん送ってくれた。メールで「10億ドル企業を作れるはずだ」と書き、数ヶ月後には「100億かも」——と彼が覚えている。

EN

Barely. It was it was a little bit profitable. But the I mean, back then, nowadays like ARR ramps are astounding. But and as are valuations. Back then, to build a business in mobile and have it go from zero to $12 million run rate in that short amount of time was really phenomenal. And so, once we did that, it was easy to recruit investors in November. I mean, it's like, okay, looks pretty good. It It was more around for a few folks that I just known for a while and a few other folks that I thought could be helpful on to the business and to mark my own dollars in. And so, we did that round in November, and then we didn't raise another round. So, back to the board point and we did raise another round, but it was more of liquidity later, but back to the board point, I didn't have a board in this business until 2018. So, from

フォロウギ#58

数ヶ月後ではないが、そういうメールはある。

EN

Six years later.

センラ#59

Raphとは?

EN

Yeah, six years later when we actually closed a round from KKR and I can talk about like what went into actually

フォロウギ#60

高校中退。16歳で前の会社から一緒に働き始めた。今年34になるから、約18年。弟のような共同創業者で、会社の人間としてすごく差をつけた。ここでは最初のビジネス採用の一人。

EN

Oh, this story's wild. We're definitely getting to that.

センラ#61

彼が先に君を雇おうとした?

EN

It was It was a messy path. I would say like pros and cons to to not having a board. The pro was I made every decision on every shareholder certificate. It was literally my signature as the president, the vice president, the everything. It was like the the I think I had three or four signatures on all the share certificates. And so, I can make all the decisions for the business and it all flowed through me. That was nice. On the other hand, I ended up making some mistakes, especially in the capital markets and raising capital, that I wouldn't have made if I had had a board and and people who had understood a lot of things that I ended up navigating poorly.

フォロウギ#62

19歳のころ、自分の事業に僕を雇おうとして、顔で笑った。「その事業はクソだ。でも君にはいい機会がある」と振り向けると、彼は断ってこっちに来た。最初の6ヶ月はオフィスに住んだ。比喩じゃない。二段ベッドがあった。シャワーは汚いのしかなかった気がする。ある夜誰かが侵入しようとしてビデオに映り、6ヶ月で「出る」となった。でも24時間だった。

彼の超能力はディールを誰より上手く結ぶことと、15年前の会話の一言一句を覚えていること。僕は昨日何をしたかも忘れる。BDを走り、いまはマーケティングとグロース。

採用でいつも見たのは「肩にチップがある人」——押す理由がある人だ。自分にもあった。高校中退でビッグテックには入れないが、売りとカリスマは誰より上。入れて権限を与えれば、他より押す。それが公式だった。

EN

Can you give us some examples of that?

センラ#63

いま彼は億万長者でも、日曜の夜11時に電話すれば出る。まだすごくハングリー。

EN

Yeah, so in 2014 and 15, we were growing really, really quickly. And we got to a point where So, 2015, call it some period, I think it was about three three and a half years post launch, we're going to do about $50 million of EBITDA roughly. But that's pretty big ramp up. Again, back then like this is I was I was wealthy enough, but the rest of my team wasn't. And so, this is a really quick ramp up for a business. We don't have a lot of people. We were just growing really quickly. We started getting interest from from other tech companies to buy us. And so, got as big as an offer for $600 million cash. From Snapchat? No, I'm not talking about the company that's on the other side. But there's a reason why I like running my own business. It's tough for me to imagine being at the company on the other side and and uh the companies that we were engaging with might have been even tougher for me to imagine being an employee at. And so, the at this point in time, the business was really all roads ran through me as a a really hands-on executive. And the challenge with that was we were still growing 100% year over year. So, when you're going and and trying to do a deal and someone presents a term sheet for towards you and goes, "Look, there's a lot of zeros, $600 million." You're like, "It is a big number." Like, all cash, like life-changing for a lot of people on the team. The problem is between the time of seeing that number and getting to a deal, your business has grown a ton. And I'm a finance person. I would have had a number I would have done the deal at, but it was not that number. And so, I ended up walking away from it.

フォロウギ#64

広告主と話して「1日あと5000ドル」と言うと、「億万長者だろ、何やってる」と笑われる。みんなそうだ。肩のチップがないと「金があるからその電話は取らない」になりやすい。

EN

What was the number?

センラ#65

Rupert Murdochのドキュメンタリーで、父のメディア会社を壊した人への恨みを45年持っている、とあった。僕はそんな恨みをほぼ50年持てない。

EN

It was quite a bit higher. It had to be around a billion dollars at that point in time.

フォロウギ#66

どこから来たかを忘れないことがある。家族はイランを追われた。僕は4歳でここに来た。父は国内で最も成功した不動産デベロッパーのひとつを持ち、何千人も雇い、裕福だった。国を追い出され、育つなかで父の一部が失われ、完全でなくなったのがわかった。後から強くなったが、最大の成功を証明する義務を感じた。父が全部捨てて連れてきてくれたから。今も消えない。成功の到達点は動くターゲットで、どこにも近づいていない。常に押せ。チームにも、恨みやチップ、動機がある人を探した。人生が楽すぎる人は入れなかった。

EN

Still only 400 million?

センラ#67

毎朝ちょっとパラノイアで起きるタイプに見える。成功している人の多くがそうだ。

EN

Yeah, it wasn't that much. I mean, in hindsight, obviously

フォロウギ#68

長いあいだ、毎朝起きて「統計を見る。今日破産するのか、まだうまくいっているのか」だった。今の規模では破産しないだろうが、明かりがついているか確認する。毎日同じ時間に起き、同じ朝のルーチン、同じ統計。それがエンジンになる。

EN

if you would have made that mistake?

センラ#69

この番組の定番だが、自閉症だと言われる?

EN

Yeah, it was

フォロウギ#70

妻には言われる。今は流行りでもある。自閉症かどうかはわからない。

EN

What's your market cap today?

センラ#71

そうは思わない。目を見て話せる。CoinbaseのBrian Armstrongも周囲はそう言うが、物語は上手い。

EN

140 billion currently. It would have left a lot of money on the table. Now, look, you most businesses when they sell, they lose the opportunity to keep innovating. And so, had we done it there, we may none of us would have known what this is going to become. So, it was a little fortuitous that I had enough conviction to to not take an offer in that range.

フォロウギ#72

25年前の僕は、もっと反社会的で自閉っぽく見られたはずだ。良い創業者はパターンを学び、必要なことをやる。昔はこの椅子に座っていられなかった。緊張で汗だく。大学直後の最初の面接はスーツで入り、雨に打たれたみたいにびしょ濡れで出た。コミュニケーションが仕事の要件だとわかってからは、うまくやる。自閉かどうかにかかわらず、前に進むのに必要な機能を解く人がこの役割では強い。

---

EN

to go back to the mistakes you're making uh without like where you're just saying the mi- mistakes in the capital markets that you think you could have avoided if you had a a board. We'll go back to that. But, um I want to take a small tangent because Raph, who maybe you should should explain who that is, he sent me a bunch of notes on you.

センラ#73

ミスしていたと思っていた話に戻ろう。2015年より前。

EN

Yeah.

フォロウギ#74

大きなオファーを取らなかったのはボーナスだった。投資家のボードがあれば、取る圧力があったかもしれない。自分で決められるのはよかった。短所は、断ったとチームの一部が知り、「あの金額だった」と計算できると、何かする圧力になる。

2015年後半。大学出たての頃はデリバティブトレーダーで、マルチプルと資本市場のアービトラージはわかる。当時、中国の投資家が米国テックを大量に買っていた。中国の資本市場が開いて、向こうで上場してスプレッドを取る構図だ。その投資家と話し、2016年のある時点で、バリュエーション14億ドルに対する10億ドルのPE投資を発表した。

EN

And the reason it just came to mind cuz uh he he remembers being getting an email. I think he actually sent me the the the copy from the email where you're like, "I'm pretty sure we can build a billion-dollar business here, guys."

センラ#75

誰から?

EN

Yeah.

フォロウギ#76

Orient Hontai Capital。小さなチームで、4人のいい奴らと中国で付き合い、好きだった。投資した。

EN

And then like a few months later it was like, "Maybe it could be a $10 billion business."

センラ#77

そのときの事業の状態は?

EN

It wasn't a few months later, but but uh so so he Can you describe

フォロウギ#78

まだ前年比100%成長。2016年のEBITDA 8500万を前提にしたドルの一部があった。数字を超えればフルバリュエーション、という構造。ディールの構造はキャップテーブルの大半をクリアし、僕の株はほぼ残り、ロールしたい人はロール、中国で上場、という話だった。

マンダリンは5語くらいしか話せない。中国の公開企業CEOには向いていなかった。彼らはカリスマで映ると思ったが、コミュニケーションなしは良くない。ボードがいればわかったはずのことは、トランプが第1期に当選した直後で地政学が悪化し、中国マネーが米国テックを買うことへの疑念が強かったことだ。僕はアービトラージだと思っていた。国有企業が何かも知らなかった。

クロスボーダーなので規制——CFIUS——が絡む。外国マネーが米国企業に入って国家安全保障の脅威があるかを見る委員会だ。知らなかった。ディール発表で「金持ちになった」とみんな祝い、「次は?」で「DCに行ってこの委員会を説得しろ」。NSA、CIA、FTC、DOJ……全部が部屋にいるような感じだった。

EN

Raph is though because he's a crazy story. So, he's a high school dropout. He started working for me in my old company when he was 16.

センラ#79

なぜ? 中国だから?

EN

He's worked for you or with you for over half his life.

フォロウギ#80

中国であることと、大規模データプラットフォームで大量のモバイル端末に載っていること。ソリティアの話をしても、悪用されうる、という懸念だ。DCに行って、何もわからず、弁護士を雇い、ボードはなく、みんな「10億、現金、終わり」だ。中国側は10億を中国から出すのに成功したいい奴らだったが、部屋は氷のように冷たい。ペルシャ系CEOに、CTOはロシア系エンジニアに交代した直後——ペルシャとロシアが中国企業に売る、という絵だ。最初のミーティングで「中国の人は出て」。

1年以上行ったり来たりして、弁護士も「ゴールはない」と言った。中国ディールが潰れて晒し者になりたくなかった。

EN

He's He'll turn 34 this year. So, yeah, he's worked alongside me for about 18 years. And so, I mean, I consider him little brother, like co-founder, person of the company. He's made a huge difference. And when we started here, he ended up He was the first business hire alongside this other guy at the time.

センラ#81

規制当局は具体的に何を言う?

EN

Did he try to like hire you first?

フォロウギ#82

懸念は言うが、部屋を読むと、ソリティアの話をしても全然関係ないことをメモして聞く——プラットフォームが誤った手に渡るリスクを評価している。70%売りの話だ。フルコントロールしていた会社が、部分的に国有のPEに70%支配される。彼らが何をするか。圧力でアメリカ人の利益に反することをさせるのか——当時まったく意識していなかった。相談相手がいなかった。ボードがあればこのミスはせず、時間を節約し、チームへの流動性は別手段があった。いいバリュエーション、現金10億、自分は所有を残して走り続ける、と思っただけだ。

その1年超、オプション発行も採用も凍結気味なのに事業はまた倍になり、2017年にはEBITDA1億超。ボードなら「ディールを破れ。いま14億どころじゃない」と言う。一方で、自分に賭けてくれた人への忠誠がある。「誰も賭けてくれない」と思い続けてきた側として、賭けてくれた4人を放置したくなかった。中国から10億を出してディールが潰れたら、金と人に何が起きるかもわからなかった。

だから後でほどける形のディールを探した。支配権を伴う70%のエクイティではなく、コンバーチブルノート10億にして、転換時に元本+利息を払い、彼らは10%——10%未満でコントロールなし——で乗る形にした。それで株主に10億を配当した。IPO前の流動性ラウンドのようなものだ。チームの手元に金は入ったが、巨大な転換債が残った。

いくつかのPEに電話した。ひとつがKKR。Harold Finnとチームが、机に来た「変な名前のAppLovinがPE調達」のメールを見て、名前だけなら逃げていた、と言うだろう。でも袖をまくり、良いアイデアだとわかった。伝統的な米国資本とデットで中国のコンバーチブルをクリアし、彼らに所有分を渡してキャップテーブルをきれいにする。2018年夏、初めてのボード——僕、Eduardo Vivas(Raphの兄)、Harold——の3人ボードを作った。

EN

Yeah, he tried He tried to hire me first on business he was working on and I laughed in his face.

センラ#83

共同創業者は当時どこに? 3人で創業?

EN

Oh, but he was how old?

フォロウギ#84

そう。Scott John Kristynakは技術担当で、当初CTO。良いチームを作った。良いマネージャーの仕事は良いチームを作り、最終的に自分と入れ替わることだ。最初のエンジニアの一人 Basil(Basel)Shikin がすごく良くなり、2016年にCTOを彼に替え、Johnはそのころ退いていった。

EN

At that point in time.

センラ#85

会社から完全に?

EN

like 19, right?

フォロウギ#86

そう。リーダーに安住しがちだが、下の方がその役割に今より合うなら、下は辞めて新事業を始める。全員と会社に対して、常に最適の人を最適の席に置く義務がある。つい先週も、Baselが別役割に移り、彼が雇った人がCTOになる、と同じことがまた起きた。10年いるVPの下に2年の方がはるかに上手く、席を譲らないと新鮮な血が抜け、長期在籍者が重要席を占め、アップサイドがあるのに人が育たない——憂鬱になる。トレードする。

もう一人の共同創業者はプロダクト寄りで、3〜4年前に退いた。プライベートの高速成長・小チームから、上場・大きな会社へ移行すると、次の章に要る人のタイプが変わる。50〜100人規模で構造なしにハッスルする人から、公開企業よりは薄いがスタートアップよりは厚い構造でもハッスルできる人へ、チーム全体で移った。

EN

Yeah, he was 19.

センラ#87

タレント維持とAプレイヤーの文化について面白い考えがたくさんある。EBITの98%がだいたい400人から、みたいな話もある。

EN

And then you tell him, "I'm not working for you. Your business is shit." Is the story I've heard.

フォロウギ#88

そう。

EN

And I I I spun it to like I got a good opportunity for you though. And he shut that down and came over. But he was great. He For the first 6 months, he lived in the office. And and uh we're in downtown Palo Alto.

センラ#89

後で聞く。まず、ほぼ6億で売ろうとしていた事業で、中国には10億で10%?

EN

But that's not like a That's not a figure of speech. He had a bunk bed.

フォロウギ#90

コンバーチブル後から今日までで、だいたい50〜55%薄まった計算だ。いま所有は約12%。当時は約24〜25%。その薄まりが、中国への10%、KKR、上場までの全部。

EN

Yeah, he actually lived in the office. He had a bunk bed in the office. I can't remember where he showered. I think we had a dingy shower. But um one day someone tried to break into the office at middle of the night. We caught him on video. It's hilarious. But I think that was it. He was like, "All right, it's been 6 months. I got to get the hell out of here." But he He was 24 hours a day. What his super skill is is he strikes deals and does them better than almost anyone. And he can sit down with you and have a conversation today. 15 years later, he'll remember every word in that conversation. I forget what I did yesterday. So, as the counterpart to me, he was great. Um and so, he ran our business development efforts and he was really helpful in it. Now, he runs marketing and growth marketing. That's why you guys ended up intersecting. But we ended up with like just this group of people that almost were outcasts from other places. Like one of the things that I always looked at for hiring were find people who have a chip on their shoulder. Someone who has a reason to push hard. And I had that myself. And and we can talk about that if you care, but that was always what drove me. I had a reason to be driven. And so I always looked for that in people. And you have this kid who's high school dropout like anywhere else, he's not getting the job at Big Tech. Like but skills that smoke everyone else when it comes to selling, the charisma through the roof. And so seeing that, I knew bring that person in, empower that person, and they're going to be hungry and they're going to push and they're going to push harder than others. And that was a formula we always had.

センラ#91

KKRとIPOの間は何もない?

EN

today, like so he goes up from high school dropout to a billionaire today. Dude, I can call him at like 11:00 at night on a Sunday and he'll pick up. Like he's he's still very, very hungry and pushing it.

フォロウギ#92

ない。IPOはだいたい200億バリュエーションで20億調達、フロートは7〜8%くらい。低すぎて、振り返るとミス。

EN

talks to advertisers. He'll laugh and go, "Dude, you asked me for an extra 5,000 bucks a day. You're a … billionaire. What are you doing?" Yeah. Like all of us are that way. I mean, but but it again it comes back to you got to have a chip on your shoulder. It has to be a reason cuz it's easy to go, "I have a lot of money. I'm not going to take that conversation." But that's not how it is.

センラ#93

なぜ?

EN

Uh I was watching this new documentary on Rupert Murdoch and like the crazy … that's going on in his family. It's on Netflix right now. And they said that he had a boulder on his shoulder. And he he's been holding a grudge for like 45 years about this these people that destroyed his dad's media company. And I was like, I didn't even I don't even have any grudges, but I don't know if I can hold one for almost five decades.

フォロウギ#94

もっとフロートを出してボラを下げたい。僕らはボラが激しい側だ。バリュエーションを押し上げてラウンドをSeries D/Eのようにすると、僕らの場合は実質Series BのようなIPOになり、フロートが少なく、長期のブルーチップが付きにくく、公開市場でボラの多い時期にサインアップする——実際そうなった。

EN

Yeah, sometimes you don't forget where you came from. I mean, mine was that my family was uprooted out of Iran. We ended up coming here when I was four, but my dad had one of the the most successful real estate development companies in the country. So he he was loaded, very successful back in the day. Thousands of people worked for him. And we got booted out of our country. And so growing up, I could just tell a a big part of him was lost and he wasn't whole. And so I felt this obligation and I like it didn't kick in for me till later, but I always felt this obligation that I got to do whatever I can to to prove that I can have the utmost success because he gave everything up to to bring us over. So that's never gone away. I I think to this day, one, this notion that like you can reach a point of success like I don't know what that is, but it's a moving target. Nowhere near that moving target. And two, always got to be pressing. And so, I always looked for people on the team that had some reason to have a grudge or a chip on their shoulder, something that would motivate them. If they didn't have that and life was really easy, those were not people that I was going to introduce to the company.

センラ#95

KKRのあと、また売却を探した? それともIPOが本線?

EN

Yeah, you strike me as a person that kind of wakes up maybe paranoid a little every day. Yeah, I feel like most successful people are.

フォロウギ#96

それ以降、買収の議論は一度もないと思う。10億の配当でみんな、僕も含め、始める前よりはるかに豊かになった。流動性を心配する段階は過ぎ、長く走ってどこまで行けるかを見るモードだった。

EN

I mean, there is a long time and I don't know I don't know if I'm past this yet or not, probably not. I'd wake up every morning and go I got to check stats. Are we going bankrupt today or are we still doing well? And it's like certainly like at the scale we're operating at we're probably not going bankrupt, but like you check to make sure like lights are still on, things are still working. And like I have I wake up at the exact same time every day. I do the same exact morning routine every day. Check the same stats. And it just gets me going.

センラ#97

それが「6年で10億企業と思っていたのが、100億かも」のメールの頃?

EN

This is a recurring theme on this show. Does anybody accuse you of being autistic?

フォロウギ#98

上がるたびに、信じられるなら口に出す。2012〜14年の頃、2014年に「10億ドル企業になる」と言った。中国ラウンド発表のオールハンズで「100億になる。やることだ」と言った。上場後、初日に15%くらい落ちて「シートベルトしろ。ボラだ。投資家は物語を好きじゃない。僕は売りが下手」。でも「クソだと言われると燃える。証明する。1000億への計画がある」と出した。

EN

Sometimes. I mean, my wife will accuse me of that, but I think you know, it's it's it's more popular these days. I I don't know that I'm autistic, but

センラ#99

それはいつ?

EN

No, I don't think you are.

フォロウギ#100

IPO後。株が15%落ちたあと、部隊を鼓舞したし、信じてもいた。時価が250億前後なら4倍。38億を切ったときは「25倍は遠い」と思った。いまキーメンバー向けのコンプに、トリガーとして兆ドル(trillion)までのプランがある。信じている未来をチームに言い、自分がまだ全力で働いていれば、ついてきて良い場所に着く——過去がそうだった。

EN

Yeah, I can I can I can look people in the eye when I talk most of the time, but

センラ#101

KKRからIPOまで約3年半?

EN

Although I didn't think Brian Armstrong from Coinbase was and everybody that knows him is like sending me messages like he's definitely autistic, but he's very he's got a great personality, he's a great storyteller, like he's not the like super antisocial autistic.

フォロウギ#102

2018クローズ、2021年4月上場。約3年半。

EN

Yeah, I mean, if you knew me 25 years ago, I'd be deemed a lot more antisocial autistic than I am. I've learned I I think a lot of good founders can follow patterns and figure out how to do things. There was a point in time where I never would have sat in this chair across from me. I would have been too nervous. I'd be dripping sweat. Like my first interview out of college, I walked in with a suit and like I literally walked out drenched. It looked like I I got out of a rainstorm. So, there was a point in my life where I was a lot different when having communication as a requirement. Now I know that it's just a requirement to do on the job, so I got to do it well. And so I think a lot of people, whether it's autism or like just figuring out how to functionally be appropriate on the things that are necessary to get forward, people who do well in these types of roles have to solve that.

センラ#103

その時点の事業の中身は?

EN

Deel is how the best founders turn the world into their talent pool. I've been studying how history's greatest founders operate for a decade, and one thing they all have in common is they understand that recruiting and hiring the very best talent is your most important priority. A players recognize other A players, which is why top companies like Ramp, Shopify, 11 Labs, Uber, and DoorDash all use Deel. Many of the top founders I know have personally invested in Deel after using their product. And what they discovered is that Deel is the best company in the world at building infrastructure for global hiring. Deel will help your business hire, pay, and manage any worker anywhere in the world, so you can retain the best talent anywhere and spend the rest of your time focusing on what you do best, delivering value to your customers. The founder of 11 Labs has a great description of the value Deel can give your company. He said, "We built 11 Labs to break down language and communication barriers. With Deel enabling us to hire and support exceptional talent anywhere, we can accelerate our innovation and bring more voices, stories, and ideas to every corner of the world." Deel is trusted by over 40,000 companies and growing fast. Learn how they can help your business by going to deel.com/cendra. That is deel.com/cendra.

フォロウギ#104

かなり変わった。初期は単純なアルゴ——ソリティアとポーカーをする人はこういうゲームもする、とゲームを押し続ける。最終的にダウンロードする。2010年代にFacebookがモバイルマーケを出し、広告をコンテンツに変えたのを見て、彼らのアルゴの方がはるかに強いと推論した。ルールベースでは足りない。2018年、機械学習へのアップグレードが要るとわかった。データが要る。僕らにあるのはゲームのプレイ。Facebookはソーシャルグラフと、世界中のサイトにピクセルがあり、広告主データが予測に効く。40サイトを渡り歩いて年4回買った、というデータは興味の予測に強い。Facebookにあって僕らにない。

より強いモデルには広告主データ——モバイルゲーム内で何に課金しているか——が要る。説得できなかった。FacebookやGoogleには渡すのに、まだモデルがない小さなスタートアップには渡さない。だからゲームスタジオを買い始めた。KKR投資からIPOまでの事業は、広告事業というより14〜15のゲームスタジオを持つ姿に変わった。マーケとマネタイズのニーズを理解し、データをモデルに入れるためだ。スタジオを買い、最初のAxon 1を出した。Axonは広告モデルの名前。最初は従来型MLだが初代より大幅アップ。データを自前で飼わないと訓練できなかった。

EN

So let's go back to the mistakes you thought you were making.

センラ#105

垂直統合だ。アイデアから最初のスタジオ買収までどれくらい?

EN

Go back to 2015 before that.

フォロウギ#106

すごく速い。アイデアが出たら行く。KKRは支持してくれた。「分析しよう」ではなく「Adamのアイデアに乗れ」。ほぼ即、PeopleFunを買い、内製スタジオも育てた。論理は健全だった。大きなボードが恐れそうなのは、ゲーム会社を支えるプラットフォームが、自分で競合する側になることだ。信頼と評判は一瞬で壊れる。僕は透明派だが、「配信側が自社ブランドを持つAmazon型」に見える。目的はデータのブランドだ、と言っても、クライアントから見ると自分の最善が守られているとは信じにくい。

EN

the the not taking the big offer was a a bonus for me because if we had a board of investors, I may have faced pressure to take the offer. And so being able to make my own call on that was like, "Okay, this is great. Like I don't actually have that pressure." The the downside was the second you say no and parts of your team know that you turned down that much money and they can run the math and go, "We would have made this much." You end up with under of pressure to do something. So This was late 2015. And I used to be a finance background. I was a derivatives trader out of school. Um and so I understood multiples. I understood capital market arbitrage. And late 2015, what was going on in the markets then and was that you had China investors buying a lot of US tech companies because the Chinese capital markets had just opened up. You know the the market over there, their NASDAQ and A shares back then was dripping. And so you had these investors coming in and saying, "I'm going to deploy China capital into US company and then go list that US company in China, make a big spread, call it a day." Um and so I went out to talk to some of these investors and ended up announcing, I think it it was some it was some point in 2016. I don't remember the exact date, but I announced a private equity investment in the company of billion dollars over 1.4 billion dollar valuation. So fast forward less than a year, company is growing

センラ#107

ゲームの経済よりデータが欲しかった——広告の方がゲームより良い、と伝えられた?

EN

Who's this? Uh investment coming from? The Chinese company?

フォロウギ#108

時間とともにそうなった。一時、空間の信頼を失った。自社ゲームが速く伸び、データをモデルに入れ、プラットフォーム上の早期採用が自社ゲームだった。競合だと見て離れる会社もいた。「論理的だ。賭けた。生きる」。一方、中小はどんな勝ちでも要る。「データ共有して事業を伸ばして」と来る。

EN

and Orient Hontai Capital

センラ#109

自社ゲームで効いているならプロダクトは動いている、と思うが、データを渡すと競合を強くするのでは?

EN

Okay.

フォロウギ#110

それが懸念だ。クローンされるのでは、と。相手を知る必要がある。興味深いデータポイントは、子供の頃はゲームをやったが、大人になってゲームが嫌いだということだ。時間を食い、中毒で、上手くなる忍耐がなく、下手なことが嫌いでやらない。開発者と話すと、「この会社は僕を通して動くが、ゲーム開発者にはなれない。コンテンツに興味がなく、データと配信だ」とわかる。手段だ、と。

ミスは、コミュニティに直接・明示的に「ゲーム事業に入る。理由はこうだ」と伝えなかったことだ。自分で知って悪意があると思われた。食事で座って質問し放題にすると克服できた。クライアントの事業が良くなれば僕らも健全だ、が信念だ。モバイルゲームはVCが少なくブートストラップが多く、創業者がマーケもやる。米国外——欧州・アジア——が多い。パートナーが要る。戦略のコミュニケーション不足で信頼を壊し、座って話せば戻る。Axon 1(2020)から今まで、彼らははるかに大きなスケールで使い、ゲーム領域で支配的になった。マーケティング費の50%超が戻ってくるかもしれない。リターンが史上最高だから事業が健全に伸び、技術への投資と信頼の組み合わせがここ数年で良い場所に来た。

EN

Um was the name. And this was like like a small team. I met four guys, really good guys. Like we hung out in China, got along. So I liked them. And they came in, they invested.

センラ#111

スタジオは売った?

EN

What's the state of the business at the time they're investing?

フォロウギ#112

ゲームが得意にならないと揃えた。全広告主が正しくモデルに接続する段階まで来て、スタジオは不要・頭痛・散漫だとわかった。広告コアは約400人なのに、スタジオは1500人超で世界に分散。回せない、DNAじゃない、見ていない。長い付き合いの創業者の会社——英国のtriple.dot——にまとめて売った。15を一つずつ利益最大化で売るのではなく、まとめて取ってもらえる価格にした。だいたい1年前。所有は最初から最後まで約5年。

---

EN

We're still growing 100% year over year. Um I think at the time part of the dollars were based on us doing 85 million of EBITDA in 2016.

センラ#113

従業員対時価総額の比率が狂気だ。キャッシュフロー年数十億の1500億ドル企業を、どう回す?

EN

Okay.

フォロウギ#114

いまはもっと多い。ランレートで、直近発表だと四半期だけで現金約13億——年で50億超、コアは350〜400人。

EN

And so like deal structure was like if you can knock this if you can beat this number you get full value. Now the structure of the deal was a little interesting. It was to clear out most of the cap table. My shares mostly would remain. Anyone else who wanted to roll could. And then we take the company public in China. Now I don't speak I speak maybe five words in Mandarin. I wouldn't have been a great China public company CEO. They thought I'd be charismatic and appear well, but no communication probably not a good idea. The other thing I didn't know, and this is where a board would have been helpful, is at the same time President Trump had just been elected to his first term and so this was actually fall to going into winter of 2016. So President Trump had just been elected geopolitical climate was getting worse. China US there were there was a lot of suspicion around like what are these Chinese companies doing buying all these US tech companies. I felt like it's just an arbitrage like whatever people are trying to make money who cares. I didn't even know what a state-owned business was at the point in time. It turned out like in in you announced the deal you still have to go through regulatory and we're doing a cross-border transaction and the regulatory body that governs cross-border transactions is committee called CFIUS and so this committee exists to go is there a national security threat with this foreign money coming into the US to invest in a US company? I didn't know what this was but announced this deal people think you got rich everyone's congratulating you and I'm like okay well what do I have to do now? What's the next step? It turned out the next step was you got to go to DC and you got to convince this committee this made up of like like every government agency I there was NSA CIA FTC DOJ like everyone was in the room.

センラ#115

どうやる?

EN

Why? Is it just because it's China?

フォロウギ#116

いくつか。2022年に株が92%落ちたとき——

EN

Because it's China for one and because we at this point we're large-scale data platform we're sitting on a lot of mobile devices and I'm like okay what do we know about people people like solitaire games like who cares but like I guess like the belief would be that if someone nefarious gets access to your technology they could do worse things with the data than than what we had access to. And so that that was the general idea but I I marched into to DC and I'm like I don't know what I'm doing I hired a law firm no board no one's no one's guiding me everyone's just like oh billion dollars get the cash call it a day. On the other side I got this Chinese investor who these poor guys that I was working with seemed like great guys they get their billion dollars out of China so I'm like okay you're doing your part like money's ready to go but we we walked into DC and it was just icy it was it was cold I mean, first of all, like I'm a Persian CEO and then at the time CTO had just transitioned to be one of our first engineers, this Russian guy. So, I'm like, this Persian guy and this Russian guy

センラ#117

なぜ落ちた? 21年上場、22年崩壊、時価40億未満。

EN

are selling to a Chinese company.

フォロウギ#118

チームを残す必要があった。ただし「全員」じゃない。機能として代替可能な人と、長期で絶対に要る人がいる。上場前は「用務員にも株を」タイプだった。上場後のボラで「エクイティが下がった。コンプが下がった。補填するのか」と聞かれた。上がったときに株を返すのか、と。返さない。嬉しいだけだ。

4年グラントを全員に配っていた。年15万のうちエクイティ5万が90%落ちたら年10.5万で家賃が厳しい人もいる。一方、キーエンジニアで年100万、うち株60万が10万になってもまだ50万——乗り切れる。低コンプの人は現金に寄せ、株は誰でも買えるが強制しない。キー——主にプロダクトとエンジニア。ブランド広告はやりたくないので営業・ビジネスは大きくない。コアが良いプロダクトとモデルを作ればスケールする。エクイティを渡すのは100人強に絞った。

誰が長期で一緒にいるかを強制的に明確にした。事業はもっと人を抱えていたがコアに絞った。AプレイヤーはB・C・Dと働きたがらない。Aの周りが悪いとAが燃え尽きて辞める。2023〜24、事業は前年比ほぼ100%成長なのに約40%を解雇した。Aに蒸留し、新鮮な血、キーロール、エクイティとアップサイド、その周りのサポート、リーンな組織にした。

EN

And selling to a Chinese company with like four Chinese investors. I'm like, that this this might not go off. Um and so, the first meeting was like, the government agency still was like, get the Chinese people out of the room. Like, it's not like just you guys. Um and so, we we go through this process and I just realized like this took the better part of I think it was over a full year, actually, of just going back and forth with DC like it trying to to navigate this process with our lawyers and they're just like, this is just not going to work. There's no way we're going to get over the finish line. And I don't I don't want to be made an example of with trying to do a Chinese deal getting blown up.

センラ#119

22から24にかけての話で、広告事業だけ? SaaSなど他は除く?

EN

What are the regulators regulators specifically saying to you? Like, what are are they voicing What are their concerns that they're voicing? Are they like trying to couch it

フォロウギ#120

今は350〜400。当時はだいたい倍近かった。GiovanniがCTOになり、2022年11月入社。BaselとAxon 2を作り、Giovanniがリードしてチームを走らせる。彼が「なぜこの人がいる? このチームは? このプロセスは? なぜこのVPで、下の方が上手いのに?」と毎日問い詰めた。最初はプロセスもミーティングもほぼゼロにしたかったが、10年以上と上場でプロセスと人が増える。

新鮮でハングリーで、僕よりずっとIQが高い人が問い続ける。答えなければ彼や同クラスを失う。文化の哲学を変えた。創業時の文化が永遠に正しいと思い込むのは間違い。世界と技術が変わり、会社も進化する。毎時点で今日の世界に合わせて文化を考え直す。LLMと自動化が来て、キラー貢献者とサポートはあるが、ヘッドカウント対売上は低くてもプロセスは膨らんでいた。

「今日創業するなら何で始めるか。1〜2年で自動化される役割は? Giovanniが周りを見て興奮できるか」——カットを始め、リーンにし、Aが汚染されないようにした。長くいる人も、LLMで消える役割も動かした。行き止まりの仕事に置かない。プロと個人の成長が本人のためなら、なければ次へ。リーンにしてうまくいった。

EN

voice concerns, but I'm pretty good at reading a room and it's like, when you give them a lot of information and you're talking about people playing solitaire and on the other side they're taking a bunch of notes and really asking you questions that have nothing to do with solitaire, you start realizing what the concern is, which is in the wrong hands as a platform that's as large as ours, it could become something that's that's risky. And so,

センラ#121

Aだけのチームに人数の上限はある?

EN

But they'll never specifically say it.

フォロウギ#122

二種類いる。公開企業として明かりを点けるサポートは必須で、全員がイノベーターのAではない。それを脇に置く。コア——特にプロダクトとエンジニア。エンジニアはプロダクト人材でもあるのでプロダクト組織は薄い。Aを揃えるにはAを見抜ける人が要る。GiovanniはA+で、A以外に我慢がない。「このレベルか、チームにいないか」。リーダーがそうなら、キーロールに不適格を置かない。Aは切らない。AI時代、リーンなチームでAIを使える最良の人と普通の人の差は線形じゃない。はるかに大きい。リーダーがA/A+で下のAを見て、切る覚悟がある。スマートだけでは足りず、作りたいものに本気でなければ、良い退職金で次へ。

EN

They won't specifically say that. That's what they're That's what they're trying to assess it for. Now, we're they're we're selling 70%, right? This is a company that I've had full control over. And so, you have this Chinese party that turned out to be a state-owned partially state-owned private equity firm fund. Again, I didn't even know what that meant, but turns out to be partially state-owned. And they're going to invest this money into us and they're going to get 70% control. So, what could they do with that control?

センラ#123

今でも、誰かを雇うなら直接君の承認が要る?

EN

Whatever they want to.

フォロウギ#124

面白い話だ。HRに入れ替わりがあり、僕がHRを見た。カット前に膨らんでいて、「なぜヘッドカウントが増える?」——誰かが辞めると自動でバックフィルの求人が上がる。だから減らず、他の採用もあるから増えるだけ。求人が何百。新人を採用して価値を出すまで6〜12ヶ月。その人がいない期間を生き延びているなら、本当に要るのか。でも「ヘッドが欲しい」と感じる。

「説得しろ」にした。年何百の採用意図が、数十になった。僕を説得するのが怖くて来ない。来たら本当に切迫している——周りが壊れそうで、働きすぎでこの1人が要る。ならやれ。

EN

I mean, it's very hard, right? Like, it's not like there are engineers on the front lines, but can they exert pressure on an American to then go against Americans' best interests? That that becomes a concern, right? Like what can they do for that kind of money? These are things I had no I wasn't aware of whatsoever. So again, I had no one to turn to. There's no one to talk to about this stuff. So not having a board, you're going through this and I'm like, okay, well, had I had a board, I probably wouldn't have made this mistake. I would have saved some time. The and there would have been plenty of alternative means to get liquidity to the team. I just thought good valuation, good number, a billion dollars cash, and I still retain my own ownership and run the business forward. So, we go forward and the business was doing really well. For this one-year plus period, we couldn't issue options. Everything was frozen. We weren't really hiring people. Yet the business doubled again. We got to like I think it was 2017, you're talking over well over a hundred million dollars. We would have thought we cleared whatever numbers we told them. Everything's going well. If I had a board, they'd say, "Just rip up the deal. This is a mess. Walk away. Your business is worth much more than the 1.4 billion-dollar valuation now." On the other hand, I've always felt like if anyone bets on me, especially as someone who's like just always had a you're not going to bet on me mentality, and I'm going to prove to you how wrong you are. I feel like if anyone bets on me, takes an investment in our business, got to be loyal the other way. And so, these four guys bet on us. They took made an investment and or thought they were making an investment. And so, I didn't want to leave them hanging. I also didn't know what happens to a group of of people from China that take a billion dollars offshore and then the deal blows up. Like what happens to the money? What happens to those people? So, I didn't want to deal with any of that. And so, I ended up finding a way to get them a deal that then we could unravel later. We ended up doing it so that instead of their billion dollars being in an equity investment for 70% where they would get control, which is what triggers this scrutiny, we ended up pivoting it to a billion-dollar convertible note that if on conversion we pay them back the money plus interest, they would get 10% of the business. So under 10% there's no control, and then they can come along for the ride. We did that deal, we dividend out a billion dollars to shareholders. Think of it as like a liquidity round pre-IPO. I knew that like, okay, we're going to get the the money in my team's hands. No one's selling their shares, but now I'm stuck with this massive convertible debt that I got to go unwrap. And so I called a few private equity funds. One of them was KKR, and that's what brought me to KKR as an investor in the business. Their partner Harold Finn and his team looked at the the company. I think he would tell you, you got an email that came across his desk that said this goofy name company AppLovin is looking to raise some private equity. My evaluation looks good. And he would have said like, just based on the name, I wouldn't have invested and ran for the other way. But his team rolled up their sleeves, they started looking, and then Harold rolled up his sleeves, started looking, and we all realized this is the good idea. Bring in traditional US capital, get some debt, clear out the Chinese convertible debt, give them their ownership, off we go, clean cap table. And so, we did that. I believe it was in the summer of 2018, and we had to build the first board that I'd ever been part of. We had myself, Eduardo Vivas, who's Raph's older brother, and then Harold Sham in this three-person board.

センラ#125

誰もが同意する「会社の価値はタレントのレベル」を、実際にAをB・Cで囲んで疏外しないという難しい仕事までやるのが珍しい。Jobsは一生でAだけの会社はPixarだけ、400人で、当時のAppleは3000で全員Aは無理、と言った。同意だけで実行しない。

EN

Where are your co-founders at this time? Cuz AppLovin had It was founded by three people, correct?

フォロウギ#126

今のソフト世界の問題は膨らんだ労働力と平凡なタレントだ。半分切っても全員が非Aなら、半分の平凡が残る。ラッキーだったのは、ずっとチームのチアリーダーで、長く自分で意思決定しプロダクトを回しつつ、自分より賢い人で囲もうとしたことだ。IQは並で特別じゃない。パターンと問題がよく見え、解を提案して進み、間違えたら速くピボットする。採用では常に自分をアップグレードする——自分より上手い人を取る。Raphの話のように、会話を覚えない僕には、その機能で例外的な人が要る。謙虚にアップグレードし、チアリーダーで、成功に必要なものを渡す。創業から変わらない。

多くの事業ではエグゼクティブが入り、雇ってチームを積み、8人のコアがプロフェッショナル・マネージャーに囲まれ、膨らみ、好きだった良い人が幻滅して抜け、惨めな場所が残る。ハンズオンで制御してきたから起きなかった。今もCROがいない。広告事業でCROがいないのは珍しい。COOもいない。Cスイートは僕、CFO、CTO、ジェネラルカウンセル。それだけ。Aを見つけたら疏外せず、育って貢献できる環境を渡す。スマートな人はスマートな人と働くのが好きだ。

EN

Yeah. So, one of my co-founders, Scott John Kristynak, was in charge of technology.

センラ#127

君の事業の作り方を数語にすると、無慈悲な効率とハイパーコンピテンスだ。VCスタートアップ界の「金を投げて何千人も雇う」への反応? それとも昔から?

EN

He's CTO or not?

フォロウギ#128

VCにノーと言われたのはラッキーかも。YESだと、ラウンドを使え、もっと雇え、次のレベルへ、利益は気にするな、となる。僕らは利益を気にした。金を追加で入れたくなかった。十分あるが無限ではない。黒字にしたく、黒字になってから気づいたのは、雇う人と働くとき、いつでも「この惑星で最良の役割」だと信じさせたいことだ。最良=プロと個人の巨大な成長機会。情熱を持てる仕事は成長を与える。それを奪うなら辞めるべきで、安住して残るなら理論上は押し出すべきだ。

スマートな人に伸びる余地を渡すにはリーンで効率的でいる必要がある。でないとプロセスと他人に埋まる。スマートな人の共通特徴は、他人とやりたくない——ただ事を終わらせたい。途中で人が詰まると燃え尽きる。そのパターン認識で、スマートな人に所有と運用の余地を渡す信念からブレなかった。

EN

He was CTO originally. So, he built a really good team. So, I would say like, a really good manager's job is build a really good team and eventually get replaced. And the first engineer, that was the lead engineer at the time, this guy Basil Shikin, um just became so good that in 2016, we swapped him into the CTO role, um and John then at that point was transitioning out.

センラ#129

同じ頃やその後の、もっと資金のある競合はどうなった? Michael Dellは同じことをやる人が街に何千もいて、より資金があったのに全員を吸った、と言っていた。

EN

Out of the company completely?

フォロウギ#130

市場は大きい。1社はいま公開、他もいずれ公開だろうが、スケールでは誰も近くにいない。VCが付けたプライベートは、安く売ったか消えた。よくあるのは金を集め、人を大量に雇る。途中で「機会があるのになぜ雇わない?」と言われる。いまゲーム外へ広げ、あらゆる会社が広告できるプラットフォームにしている。普通は「セールスフォースとCROを雇え」になる。AIツールで、本当に良い1人が10、20、50、100人分になりうる。先に雇うより、エージェントを使って昔の大勢分の仕事をやる個人を見つける。高確信でこれに固執した。機会に合わせて雇ってIQを薄め、Aを失うと解けない——終わりだ。

EN

Yeah. And so like, I I've always had this belief that like, a lot of people get complacent with the leaders that they have, and those leaders, if they're not at that moment as good for the role as someone underneath, that person underneath is going to leave and start a new business. So, like you owe it to every single person and you owe it to your company to make sure the best person is in the right role at any given time. And and actually this like fast-forwarding to just last week, we just announced Basel's going to going to shift gears to a different role in the business. And one of the people that he hired who's built a whole bunch of our new technology and now runs the team is going to transition to CTO. So, the same thing happened again. So, this always happens that every role I've seen is if we were complacent and said you're VP you could have been here 10 years. The person underneath who's been here 2 years is just much better and we're not getting into that VP role is going to leave. I don't want to lose the best talent. I want at any given moment the best talent in the most important roles. And so, we always do that trade and if you don't do that trade it's really depressing cuz you end up with these really long tenured people who who end up in all these really important seats, but all your fresh blood is churning despite having more upside. You could have given them room to grow and you could have transitioned. And so, we did we did the average chance we can get. And then the other co-founder was ahead of product type person and I think he transitioned out it was about 3 4 years ago. And really like what changed over time for both of them and then like I think just generally like a lot of the people that were early at the business when you go through this transitional period of like private fast growth small team to we're going to go public we're going to be a bigger company, it changes what people the types of people that you may need for that next chapter. And so, there was this transition on the team across the board where we were constantly changing those like 50 100 person type company people who really wanted to hustle with no structure to people that could actually continue to hustle with less structure than what most public companies have but more than a 50 person startup.

センラ#131

Axonが効いているかはすぐわかる——売上を売る事業、と言っていた。今の効き方と、前はどうだった?

EN

You have a very a lot of very interesting ideas on retention of talent proving every single hire. What happens like how you actually retain a culture of A players cuz you have a couple hundred people I think 98% of your EBIT comes from like 400 people now, something like that.

フォロウギ#132

プロダクトが自分で売れなければならない。営業でお願いするなら、スケールする広告ソリューションじゃない。セールスもいない。ゲーム開発者が今日つないで1000ドル使うと、確実にそれ以上を稼いだとわかる。ビジネスモデルによっては1年でトントン、5年で5000になるかもしれない。30日でトントン、半年で2000〜3000かもしれない。モデルが許す範囲で、1ドルからより多くの金を生む価格付けができる。

重要なのは、彼らがアービトラージャーになることだ。スケールの制約は銀行残高だけ。「入れたより確実に増える」とわかれば、あるだけ入れる。そのアービトラージが崩れるまでスケールを取る。eコマースも同じ——ゲームをプレイする10億超の人に、ゲーム広告だけでなくeコマース広告を見せる。同じアプローチ。ドルに対する売上を渡し、測定可能にし、儲かっているとわからせる。スケールする。パフォーマンスマーケをそう定義する人は少なく、実行した会社も少ない。Facebookは素晴らしくやった。Googleの一部も。僕らはそこに全振りだ。儲かれば使う。

モデルの反復は、LLMの消費者UIのようなバージョン番号ではない。リサーチサイエンティストが改善するたびに、広告主のROAS目標でより多く使わせ、より多く売上とリーチを取れるか。精度が上がれば出力が良くなり、ユースケースが増える。実験フレームワークで価値を抽出する——広告主が豊かになること。彼らが豊かになれば僕らも豊かになる。

EN

Yeah. right.

センラ#133

グラフの変曲点を見た。

EN

So, we'll get to that in in 1 second, but I need to understand that. So, go back you almost you almost sold the business for $600 million, right? Again, we're talking about a business

フォロウギ#134

事業が大きく曲がったのは2023年4月ごろ、Axon 2のロールアウトだ。最初はルールベース。Axon 1は従来型ML。Axon 2はディープラーニングで、広告主への価値予測がはるかに複雑で高い。頭の中の価値提案——広告主をアービトラージ・マーケターにする——は、最初は不可能、Axon 1では金と学習と手作業が要る、Axon 2ではつないですぐ使い、使って売上とスプレッド——まで来た。スケーラブルなプロダクトがあると事業は大きく伸びる。

株が9〜11ドルだったころから、ピークはたぶん半年前に750ドル前後。バリュエーションは40億未満からピーク2500億。その後下がったが事業はまだすごく良い。時価総額の伸びはファンダメンタルズが並走したからで、Axon 2が広告主にその価値を届けたから過去3年の成長は凄まじい。

EN

almost I I know.

センラ#135

キー人材の報酬トリガーが兆ドルまで、と聞くとクレイジーに聞こえるが、AIの追い風と、Aだけのリーンなチームで、1人が4000人分をやる世界なら違う。直近3〜4年でAIがどう事業運営を変えた?

EN

Or you were you were $400 million off.

フォロウギ#136

ツールの力はここ数ヶ月でブレークアウェイの速度だ。小さなチームがはるかに多くできる。1年半前でコードの80%超がLLM生成で、今はもっと高い。LLMと例外的エンジニアのペアで出力が増える。Bプレイヤーの世界なら1Xが2X、でも10Xは100Xになりうる。MLモデルの実験と改善の速度だけでなく、空間の研究を全部処理できる。オープンソースの研究分野で、5年前は人間が全部読んで理解できなかった。数学と研究がわかる人がLLMとペアし、抜け漏れなく実験を作る。効率のポテンシャルは巨大で、僕らはディープラーニングモデルの上にいるから技術が強くなるほど予測価値が上がる。

一方で、毎日ゲームする10億超のユーザー。米だけで成人1.5億超。浸透は高い。パワーユーザーはInstagramに1日6時間の21歳と同じではない。Candy Crushを2〜3時間やる中年で、時間がありリラックスしている。いいオーディエンス、滞在時間、広告を長く見る——平均35秒超。テレビCMがモバイルにいるようなものだ。フレームワークは良い。中身は技術と顧客が増えるほど良くなる。数年前はゲーム、ゲーム、ゲームだけ。ゲームを乗り換えない人には悪い。いまeコマースのプロダクト広告があり、いずれ世界のあらゆる事業へ。多様性が上がり、技術はすでに対応でき、エンド消費者への価値が上がる。広告がコンテンツになる、という話の希望は、地元のランドリーが、服を洗いたい人に良い広告でゲーム中に発見されること。そこまでスケールすれば、今日よりはるかに大きく、中小の経済成長に効ける。

EN

Having discussions, having discussions.

センラ#137

ゲームからeコマースへ。全セグメントに行けると本気で思う?

EN

So, how much of the business did you sell to the Chinese then? Did I understand you sold them 10% for a billion?

フォロウギ#138

ゲームする大人の数を考えれば、限界はターゲティング技術と向こう側の顧客だけだ。最大の機会は中小だ。セールスフォースがないので大企業はいずれ来るが、こちらから売らない。ゲームで成功したのは、他が手厚く支えない10人規模と一緒にやったことだ。トルコのモバイルゲーム会社が、ローンチ1年未満で約10〜15人で10億ドル弱でイグジットした。マーケとマネタイズのほぼ全部が僕らのプラットフォーム上で、ゲーム自体も革新的だったが、この種の会社を助けられると気持ちがいい。同じことを他カテゴリ——地元ランドリー、Shopifyで口紅を売る小さな店——に移せれば、すごくいい。

EN

So, so the way it worked so so actually like I sort of calculated to my ownership today.

センラ#139

DTC向け? B2Bも?

EN

Okay.

フォロウギ#140

最終的にはコンシューマーとエンタープライズもやりたい。理由がない。10億ユーザーの向こう側は大人で仕事もある。全コンシューマーカテゴリでターゲティングできれば、エンタープライズ側にも限界はないと思う。

EN

After the convertible note round, we diluted about 50 to 55%. Until today. So, if I took I own about I'd say about 12% of the company. I owned back then about 24, 25%. So, that dilution is all the dilution we've had between the convertible note conversion to the Chinese, pay out the Chinese, convert them to 10% equity, bring in KKR, eventually go public.

センラ#141

事業の回し方と同じで、この会話にも贅肉がなかった。時間をありがとう、Adam。最高だった。

EN

There's nothing between the KKR and IP and IPO. Why? So, after you do this deal with KKR, was there a chance that you were going to look for another acquire or did you know, okay, the likely outcome here is we're going to IPO? Is this when I'm reading the email where you're like, okay, I thought I think six years in maybe this could be a billion dollar company. Now you're thinking, oh, it could be a 10 billion dollar company. You said that when? So, let's say the market cap's around what? 25 In between the deal for KKR and IPO, there's what, 3 years? How how how what's time frame? What does the business consist of at that point? What does turning ads into content mean to you? How long from the idea of we were going to essentially like you're vertically integrating, right? And so now you're buying gaming studios. These gaming studios all own a bunch of different games, I assume, and they're producing. And so now you have the data of what people are doing inside the games, right? So the idea of like how long from the time you had that idea till you start buying your first gaming studio? You could make the case to them that what you actually wanted was not the economics of the game, you want the data because you still thought the advertising business was better than the gaming business. think it might have changed? Yeah, cuz if I'm them, like, I know your product works because you're using it to grow your own games. Oh, yeah. They could just clone their game. Wait, wait. You hate games? them that story, though? So, you sold off the gaming studios? So, how many years did you have to own these gaming studios? Adam is one of the most focused and intense founders I've ever met. As you're seeing during this conversation, Adam is driven to build a great product that serves his customers, relentlessly improve that product over time, and win. And that is exactly what Adam and his team has done with their advertising platform Axon. Axon connects you to over a billion potential new customers in mobile games. Axon allows you to capture undivided attention. Axon's ads are full screen videos that are watched for an average of 35 seconds. Retention that blows other ad platforms out of the water. You can launch on Axon in minutes. You set the goal and Axon achieves it. No complex setup, no expertise needed. And Axon scales quickly. They can put your ads in front of over a billion potential customers. Other businesses have seen immediate results, scale to hundreds of the of dollars to spend per day, and increase their revenue by millions. So, you want to get started quickly before all of your competitors are on Axon, and you can do that by going to axon.ai/centra. That is axon.ai/centra. Uh more than that now, quite a bit. Yeah. Run rate wise, I think Q1 we made our our I guess Q4, what we just announced in in Q1 was um I think it was 1.3 billion in cash just in a single quarter. Yeah. Couple things happened. When when we tanked in 2022, This is when the stock market stock collapsed 92%. Like in in in '21? Stock collapses in '22, that's when we we ended up under $4 billion market cap. A couple things. One is I needed to make sure I retained my team. But by team, I don't mean like everyone on the team. There's certain people that are just doing functions and are replaceable. There's certain people that are critical to a business and like you really need them in with you for the long haul. And so, what I realized was when building a business, you feel like everyone should be a shareholder. I was like one of those the founders that was like, "You're the janitor and you should get shares of my business." When we went public and we faced this volatility, we started hearing from or at least I started hearing from a lot of people like, "Man, we're way down on our equity this year. Our comp is way lower. Are you going to screw us up? Are you going to give us stock on the other side?" And super frustrating for me cuz I you think about the trade on the other end of it. Well, when the stock goes up, are you going to give your shares back to me? Like, you're not going to do that, right? You're going to be happy and you're going to get overpaid. And so, we had these four-year grants that we distributed to everyone on the team. And what I realized was it's unfair for me to judge because some people just need compensation. If you're getting paid $150,000 a year and 50,000 of it was an equity that goes down by 90%. Now you're at $105,000 a year, you might not be able to pay your bills. Like you really depended on that equity. On the other hand, if you're critical engineer and you're paid a million dollars a year and your $600,000 of stock compensation goes down 100 grand, you're still getting paid 500k a year. Like, okay. There there's enough money to make ends meet and to write out the volatility. And so, what what we did at the the low point for the team was go, some people were just going to put on cash compensation. Everybody at the company can buy shares in the business, but we don't need to force that decision for people that have lower total compensation. So, they can just go cash compensation. We then said, who are the people that are like critical to the business? Your key engineers, your key product people. Again, remember we don't want to do brand advertising whatsoever. So, like this wasn't mostly the business people or sales team cuz we don't even have much of that. This was core product and engineering. If they build a really good product, they build a really good model, we're going to be able to scale the business. And so, we took it down to a little over 100 people that we give equity to in the business. And why this was important was it forced me to go, who are the people that matter? Like, who who are who am I really in this with for the long term? And we had more people at that point in time. We we really slimmed down the business to this core group of people and the support around them. But it was important for me to get it to that point because to your point of having A players, B players don't like to work with B's, C's, D's. I mean, D's you're you're you're going to just burn your A's out. But like, anything that's worse than an A player on the team and you're distracting your best for working with those types of people and like trying to bring them up to speed it is a huge loss cuz eventually you end up churning out your A's. So, I had to not only understand who are the key people, I had to go through this period of time where in 23 either I think it was about 23 or 24, even though the business was ripping, we were firing a ton of people. We let go about 40% of the team. I think it was 24. The business grew nearly 100% year over year and we fired about 40% of the team because I wanted to distill it down to these A players. I wanted to get it to a place where fresh blood, key people, key roles, they're getting equity, they've got a lot of upside, support around them, and then we're lean and mean organization. Yeah, not talking about those things. Uh so, we're we're more, yeah. I'd say probably close to double that. And so, this guy Giovanni who just started as CTO, he joined and he he joined in November 22 and he's responsible for him and Basel that the CTO I talked about. They they basically together worked on the Axon 2 model. Really Giovanni led it and then built out the team and he runs the team. Now, the reason I raised his name is he came in and he started asking me interesting but difficult questions. Questions like, "Why do we have this person? Why do we have this team? Why do we have like these processes?" And as a business, you sort of go when I first started, I don't want any process. I have almost a no meetings rule. Like, I wanted everything to be as highly efficient as possible. But over a decade plus, you inherit some process. You go public, you're told you need these processes, you need these people. Well, you get someone in who's fresh blood, who's really hungry, insanely high IQ person. I mean, one of the smartest people I've ever met, much smarter than I am. And he's grilling me with this question every day like, "Why is this person in this role? Why is this person here?" Like, "How come this person is a VP to this person and the person underneath that person is much better than the VP? So, I kept getting these questions and I was like, "Either I'm going to have to address these questions or I'm going to risk losing talent like this person or the other people that are talented like him." In doing that, we ended up going I changed my philosophy to culture and how a company is structured. You start a business, you have culture, you have cultural values and like you hope those hold true forever. Well, the world's changing really quickly and world of technology is changing and companies are evolving. If you just consider it static where your culture from day one is going to hold true over time, I think it's a mistake. So, what we did is I I took his questions to heart and said, "At every moment in time, we've got to rethink our culture for what the world has today." And you started having LLMs at that point in time and a lot of automation was coming. And we also had a business where we had these key contributors that are killers and then we had support around them, but we had a lot of bloated process. Even we did despite the low head count to revenue because of 10 plus years of operating and just the way it was. And so, I re-challenged it and said, "If I was starting the business today, what would I start it with? What is it going to look like? What is going to be automated away in a year or two years? How am I going to get Giovanni to stop asking me why and just be excited about his role cuz everywhere he looks there's a super capable person around him." And so, I went in and just said, "We're going to kick off a whole bunch of cuts all over the board and we're going to do it in this hope of getting leaner, but more importantly keeping A people from getting contaminated so that those A people can be retained, feel like the organization is theirs, and we ended up turning over a lot of tenured people that work here for a long time. We ended up turning over people whose roles I felt like were eventually going to get automated by LLMs. And if we felt that way about their role, I wasn't going to keep them in a dead-end job. I want people to go and do whatever is best for them. And usually professional and personal growth leads into what's best for them. And if we didn't have that, we were going to move on from it. And so, we just leaned everything up and it ended up working really well for us. Yeah. You have two types of employees. You need some support functions, right? Like at the end as a public company, like not everyone can be like an A player because they're not in roles to to innovate. So, let's say There's some people that are just doing roles to keep the lights on. That's just a requirement. So, set that aside. Then you've got like your core people, especially product and engineering. We expect our engineers to be product people. So, we don't have much of a product org. So, let's call it core engineering. We expect them to be A people. Now, to to have a team of A players, you've got to be able to recognize A players. So, I know that Giovanni's an A+ player. I know that he has no tolerance for anything that's not an A player. So, he goes and he's very quick to go, "You're either cutting it or you're not on my team, but you got to be at this level or you're out." And so, putting people in these leadership roles that require that level of talent allows me to have confidence that because they're impatient, they're not going to enable talent that's not not qualified in these key roles. And so, you're always turning over people that don't meet your need and you're going as lean as possible. Now, when you have an A player, you don't want to cut those people. Those people are super powerful, especially as we go forward and you're going to need lean teams that know how to use AI. Your best people know how to use AI better than your mediocre people and it's not by a some sort of linear function. It's It's much greater than that. And so, our team is constructed to have leaders who are A players or A+ players who can see the A players underneath and who have no patience for anything that's not and are very willing to fire on site. If they feel like you're not cutting it, you're this role is not one that you're passionate about. And most of our people, because we're lean, are very smart. But sometimes being smart is not enough. You got to be really into building what we want to build. And you're not cut out for it, you you got a really good severance and you go off along your way. So this was a a funny one. And and the reason this came up is there's a period of time we went through turnover in HR and I took over the HR organization. And I saw that same period before we did the cuts, we'd become bloated. And I was like, why is the head count so high? Why does it keep going up? Well, turned out every time we let someone go or someone quit, there was automatically just a new job placement that went up. It was a job placement as a quote-unquote backfill. So I had a role, this role disappears, I'm going to automatically rehire. Which means there's no way your head count can go down and your head count's only going up cuz you're still inevitably hiring other roles, too. Now, what bugged me about that was we had hundreds of job postings. And in order to hire and train and get value out of a new employee, you're talking about 6 to 12 month process. What happens in the period of time when that person quits and you don't have that person? Why do you need to backfill if you can survive without that person, without anyone in that role for some period of time? You probably don't. But it feels better to say I want the head. And so I took it over and I said, look, convince me that you need the person. So we went from somewhere in the neighborhood of hundreds of of new hires that we're doing a year or at least attempting to do a year to in the tens cuz people were so spooked about trying to convince me that they need the role, they wouldn't come to me. And like now it's like if they come to me, I know that they're desperate. It's everything's going to break down around me. People are working so much that they're they really need this extra hire. Okay, go make it. Yeah, the the challenge I mean like you see it in the the software world today. The challenge that a lot of companies have now is bloated workforce. And the talent's mostly mediocre. So what's the solution? You can't just fire people. If you fire half the people and everyone was not an A player, you're just left with half mediocre. And so what do you actually do? Where we were lucky is that I've always felt I'm a cheerleader for for the folks on the team. And like for a long time I was making a lot of the decisions on the team and I ran product and I was hands-on, but I still try to direct people and surround myself with people smarter than me. And I I always felt like pretty reasonable IQ. I'm not exceptionally high IQ. I just see really I see patterns and problems really well and I can typically like understand what that problem is and try to propose a solution and move forward. And if I'm wrong I can pivot quickly and get to a good answer. And so maybe that was the the skill that I had but paired with that I was always humble enough to know that when you hire you always want to upgrade yourself. You want to hire better than yourself. It's something that you're not as good as. Like the raft discussion earlier where where he's exceptional at beating. Well, I don't remember the conversations I had a day ago, five days ago, five years ago. So I needed someone like that exceptional talent in what a function could upgrade me and so I can go he's much better than me at that. And so what what's important in that is that I would took the humble approach to upgrade at every level and then be the cheerleader for these people and give them whatever they need to succeed. And that hasn't changed since we started the business. And I think it's very challenging to maintain that cuz a lot of times when you're building a business, you start hiring these executives and these executives come in and they're trained to hire and hire build and they they build up these teams. And all of a sudden train leaves the station, you had this core group of eight people and all of a sudden you surrounded them with these professional managers and workforce gets bloated and those really good people that you used to love working with, they get disillusioned, they churn out and what you're left with is a miserable place. That that function has held true to a lot of businesses and because I was so controlling and hands-on, I never let it happen. To this day, we don't have a CRO. We're an ads business. I mean, anywhere you look advertising business, who doesn't have a CRO? We don't have a CRO. We don't have a COO. We have today the the C-suite is myself, the CFO, the CTO and we have a general counsel. That's it. And so, you you look at a the structure of the business, I made sure that we didn't alienate A people when we found them and I made sure that they had a perfect environment to grow, develop and contribute so that they feel good. And if you give them a place to feel good and take ownership and learn as they go, really smart people love working with really smart people. Maybe we got lucky in the the VCs telling us no cuz when VCs tell you yes, they want to put more money to work. So, it's always like you took your round, use your round, hire more people, get to the next level. Don't worry about making money. When we started, I mean, we worried about making money. I didn't want to put more money into the business. We were there I'd made enough not to worry about money, but I didn't make that much. And so, we wanted to be profitable. And then once we started getting profitable, what I realized is one of the most important things when I hire and work with someone is I want them to believe that at any given time they're in the best role for them anywhere on the planet. And so, I define best role as massive growth opportunity, professionally and personally. I think really good jobs where people become passionate and love their work give them an ability to grow. If you take that away, then they should leave. And a lot of times people get complacent and stick around, but in theory, you should just force them out. And so, what that means is to give very smart people room to grow, you got to maintain lean and efficient. If you don't, those smart people get bogged down with process, with other people. And I've also found a common characteristic of really smart people is they don't want to deal with other people. They just want to get … done. And if in the middle of them getting stuff done, they're stuck with dealing with other people, they're going to eventually get burned out in the business. And so, because of this pattern recognition, where I knew what would retain this type of talent, I never wanted to waver from the belief that we got to give very smart people plenty of room to own and operate however they want, which is in their own world, highly efficient, limited on process, and and that kept us lean. Yeah. I mean, we're in a big market, so a couple are one of them is public today. A couple more will eventually be public, but no one's even remotely close to the scale that we operate at. And And the private ones, like I said, the the ones that the VCs had funded either sold for very little or gone. And very common would be raise money, hire a bunch of people. And along the way I've always gotten the you have huge opportunity, you're doing really well, why aren't you hiring people? We have big business opportunity in front of us right now. We're as we go forward, we're really expanding the platform that we operate in outside of gaming to service companies of any kind to come advertise on our platform. Well, most people look at that and go naturally, why aren't you hiring a sales force? Why aren't you hiring a CRO? Why aren't you hitting it hard? And because the reason I don't is that now with AI tools, one person who's really good can be as powerful as 10, 20, 50, 100 people. And so, why would I go hire in advance of that versus go find those single individuals who could be really good, who understand how to deploy agents to tackle jobs that many people would have done before. And so, I've stuck to this view with high conviction cuz I've seen companies do exactly that, hiring into opportunity, dilute their IQ, and then you're stuck with a mess. And how do you unravel a mess? If you lose those A people, there's no way to do it. You're just toast. Yeah, I mean, so where we are today is and and this is where we wanted to get to, which is the product has to be good enough to sell itself. Otherwise, if we're begging for business, we don't have a good advertising solution that's scalable. We don't have the sales people to go beg for business. And so, what does that mean? Well, if you're a game developer today and you plug into our platform, you're going to spend a thousand dollars, you're going to know with certainty you made more than a thousand dollars on that spend. You may have a business model that says I want to break even on that thousand dollars in a year, and by year five I know I'm going to make five thousand dollars. And so you put the thousand in, you get a thousand back in a year, and year five you're at five thousand bucks. You may have a shorter cycle. You might break even on the thousand in 30 days, and then by six months out you made two three thousand dollars. Whatever your business model allows, you're going to be able to price into our system and generate more money from the dollar spent than than what you put into the system. Now, what's important about that approach is that they become an arbitrageur. The only constraint on them scaling in our system is the money that they have in their bank account. Because if you tell someone you're certain to make more money than what you put in, you're going to know it. Our reporting is going to tell you, otherwise you're not going to scale. They'll put as much as they have in their bank account, and we will drive as much scale as we can before that arbitrage breaks out. When we took the e-commerce business to the market, and e-commerce is us accessing a billion plus people playing games, but instead of just showing them game ads, we now show them e-commerce ads. This category started growing really quickly, too. We took the exact same approach. Give them revenue for the dollars that they spend, make it measurable, make sure that they know they're making money. They're going to scale. And if you do that and what I call a performance marketing is that, make the advertiser an arbitrageur. Very few people define performance marketing as that, and very few companies have executed Facebook's done a fantastic job of it. Google part of their business does this. This is entirely what we're focused on. If they're making money, they're going to spend. And so the models themselves, when we go and create different iterations of the models, we're not a we're not in large language models, so there's no consumer interface. We don't need to go this is Axon 2212223, but every time our our research scientists go and improve the model, the goal is can you drive more spend to the advertiser at their return on ad spend goals. So, are they making more revenue, and are they getting more reach? And the large language model gets more accurate, you you it, you're going to get a better output. And so, and obviously the the use cases go up. If we get more accurate, the prompt effectively is advertiser has all the the budget and has these goals, and can we deliver more accuracy for them? Can we get them more scale? Can we get them more spend? So, we're constantly doing iterations on our work and doing tests in our experiment framework to try to extract more value. More value is get the advertiser wealthier. They get wealthier, we get wealthier. The business itself inflected big time in I think it was April 2023 when we rolled out Axon 2. The Axon 1 model, I mean, the first version like I told you was just a rules-based system. Axon 1 was more traditional machine learning. Axon 2 was a deep learning model. Much more complex, much higher ability to create these predictions of value for advertisers. So, it went from the point of like no ability to do what this value proposition was that I had in my mind, which is turn the advertiser into arbitrage marketer, to Axon 1, they can get there, but they have to spend a lot of money, the model has to learn, they have to do a lot of manual labor to get to that point, to Axon 2, any advertiser can just plug in, get going, spend money, get revenue on the other side, and make a spread. When you have that product and it's very, very scalable, the business can grow a lot. And so, we we went it from I think it was April 2023 was when the stock was 9, 10, 11 bucks to we peaked it was probably 6 months ago or so at 750 bucks. So, the valuation went from under 4 billion to peak at 250 billion. Now, we've come down some since, but the business is still doing phenomenally well, and the only way we could have grown like that on market cap was if the business fundamentals grew as well alongside it, and it truly did. Our growth has been phenomenal over the last 3 years because this Axon 2 model was so powerful in delivering this kind of value for advertisers. I mean, for sure like the the tools that are available and I'd say like as we've all seen, there's been breakaway speed on the the potency of these technologies over the last few months. You've got small teams can do much, much more. So, it's huge. I mean, we we over I think it was about a year and a half ago, over 80% of our code was LLM written and now it's much higher than that. And so, you've got the ability for the LLM paired with really exceptional engineers to create more output. And this isn't like like if you lived in a world of B players, your 1X engineer might be 2X more efficient, but your 10X engineer might be 100X more efficient. Not only can they go out and accelerate the rate of speed that they can create experiments and improvements on machine learning models, the LLM can also process all of the research in the space. These are research-oriented sectors, open source, but five years ago, us humans couldn't have gone and read all the research and understood it. Now, very smart people who understand math and research can go, use the LLM, pair with it to create the future experiments off all the research that's out there, not not believing that they're missing anything cuz the LLM is scraping everything out. So, massive amount of potential for efficiency gains and we're built on a deep learning model. So, as these technologies get more powerful, the predictive value goes up. Now, what does that mean? Well, for us in our business on the on one hand, we have a billion users who are playing games, billion plus users who are playing games every single day around the world. In the US, over 150 million, only adults. So, penetration into the world is high. The power user who plays games is not the same 21-year-old who's on Instagram for 6 hours a day. The power user who's playing casual games, Candy Crush for 2-3 hours a day as an example, is more of like a middle-aged person who has more time and is just getting relaxation here. That's the framework. So, we have this really good audience, a billion plus, good amount of time spent, and they're willing to watch an ad for a long time. Our average ad is over 35 seconds long. It's like a television commercial on the mobile device. So, the framework is really good. What we show in that ad is only going to get much better as the technologies get better and we service more customers. And if you think about a couple years ago on our platform, we only served gaming ads. We used to take a user and say, "Game, game, game, game, game." If you weren't in the business of switching games, that's a pretty bad ad format to fill you. Well, now as we've gotten into product ads for e-commerce businesses, and it really as we go forward, it'll be for any kind of business in the world, the diversity will go up. The technology is already capable to do it, and then the value to the end consumer will go up. And that whole thing I said earlier of the ad becomes more like content, the hope we have is that we can get the local laundry discovered by someone playing a game because we serve a really good ad to someone who needs their clothes washed. If that happens and we get to that level of scale, this business is going to be much, much bigger than it is today, and the opportunity for us to really impact economic growth for businesses that are small to medium size is going to be much bigger than it is today. Is this targeted for like direct to consumer? Do you Do you think you'll be doing this B2B as well? Like the way you run the business, this conversation had no fat on it. Thanks you very much for making the time, Adam. This is awesome. I hope you enjoyed this episode. Please remember to subscribe wherever you're listening and leave a review, and make sure you listen to my other podcast Founders. For almost a decade, I've obsessively read over 400 biographies of history's greatest entrepreneurs searching for ideas that you can use in your work. Most of the guests you hear on this show, first found me through Founders.

フォロウギ#142

こちらこそ。すごく面白かった。

---

EN

Nothing else. So, like again, the IPO if I recall was about I think we sold about like 7, 8% float. So, which is really low and probably a mistake in hindsight. You want to sell more float. If If anything like you need float to like lower volatility. We've been one of the more volatile companies. If you push up valuation to where your round and IPO is just like a series D series E. For us it was our series B. Um but you end up doing the IPO and we we did $2 billion I think over $20 billion valuation. Too low a float makes you more volatile and if the valuation is lower and you can do a 10 to 15% float, you end up attracting better investors who can end up being long-term with you. If you push the valuation where you can't fill a bigger round, you end up with more volatility and less blue chip investors and you're signing up for just a much more volatile time in the public markets, which we saw. No, uh since then I don't think I've had a single acquisition discussion with anyone. Um, and by that point in time, I'd realized, well, first of all, we paid a billion dollar dividend, right? Like, and so, everyone was much wealthier than than when they started Me Too. That's more money than than I know what to do with. So, we were past the point of worrying about liquidity, and we were much more in the mode of like, okay, let's let's run this long and see what we can do at that point. Yeah, so so the you know, every level that we've gone up, I voice something aspirational, but I got to believe it to voice it. And so, when we first when when we were ripping like 2012, 13, 14, I think I first said like, we're going to be a billion dollar company in 2014. When we did the the Chinese round and we announced it, I did an all-hands where I said, look, like, we're still growing super quickly. We just did a print over a billion dollar valuation. This is going to be a 10 billion dollar business. Period. Like, now we have to do the work to get there, and we had a plan. When we went public, there was some point in time I mean, first like, we were down like 15% day one, and I was like, buckle up. One, it's going to be volatile. Investors like don't tend to like our story. Apparently, I suck at selling it to investors. But beyond that, it was look, this … motivates me. Like, I love it when people tell us we suck. They tell us we suck, let's prove them wrong. The business is going to be much bigger. We got a plan to get to a hundred billion dollars. So, I laid that out in This was after we went IPO. Stock dropped like 15%. I was trying to rally the troops, but I also believed it. I had a plan to grow the grow the business now. 25 billion. So, it's a 4x. So, like, it sounds reasonable. When we went to under 4 billion, I was like, … Like, a 25x is a lot further away. Like, people who were working on the business, we believe there's a path to a trillion dollar outcome. And we don't set time. It's just that this business can be as big as what you would need to be a trillion dollar business. And we actually just did a comp plan to some of the key people that is built on on triggers to give them much bigger payouts, but the highest one is a trillion dollars. So, at every moment in time, if I believe in a future plan and and think that it could get to that big, I want to voice that to my team cuz I know that at least my past, they've seen that if I voice something big and I'm still working my butt off, they follow and eventually we'll get to a good place. So, we closed the deal in 2018. Yeah, we went public in April 2021. So, it was about 3 and 1/2. It changed a lot, so when we first started, we grew really quickly and and we cleared over $100 million of EBITDA and we had a very simple algorithm. In advertising business, the algorithm is key if you're doing performance ads. So, our very early algorithm was, if you play a solitaire and you play a poker game, people who play solitaire and poker also play these games. And it just pushed you games. And you'd see game game game game game. Eventually, you'd download off a simple algorithm like that. What we realized um as really machine learning technologies evolved and and Facebook built this fantastic platform over the 2010s, once they got their mobile marketing platform out, they really did turn ads into content. If you have a very powerful algorithm on showing users ads that are really good for them, the consumer is going to discover products that they want to buy through the ads. And when you can do that, you can show people an endless amount of ads. They don't care about the ads that they're seeing. They're not disruptive. They're actually something that they can engage with. And what I saw was and and I inferred that their their algorithm was way more potent than what we had at the time. And now obviously the simple like rules-based algorithm wasn't going to cut it. And so, in 2018, we started realizing we've got to upgrade to machine learning capability. Other companies are doing it. Facebook's way ahead of us. In order to do that, we need data. And the data that we had were games that people were playing. The data that Facebook had, of course they had the social graph, but they're pixeled on most of the websites in the world. Advertiser data is very, very important in predicting an advertiser relevant outcome. So, by that I mean like if you're a consumer and you're on 40 different sites and you're browsing for different products and you buy four in the last year, that's a very good data set to predict what are you actually interested in. And if you don't have that visibility, which Facebook has and you don't. Facebook had we didn't. And so, if we were going to go build a much more robust model, it's very hard to just do it with solitaire and poker data and whatever else other games we had in the system. We needed advertiser data. Now, in in this case for us, it was inside mobile games, what are you actually spending on? If we could get that data, we could train a model that was a lot more powerful than what we had. What you buy is very much predictive of what you'll buy next. And so, that that was critical for us to go get our hands on that data. We were unable to convince advertisers to share that data with us. Even though they were sharing it with Facebook and Google, they would not share it with us. We were just the small startup, um even though billion-dollar company, just this other company in the space that didn't have a need for the data because we didn't yet have a model. So, in order to solve that problem early on, I went out and and we just kicked off buying some gaming studios. So, actually like the business from KKR investment to IPO ended up looking wildly different than an advertising business because we ended up with like 14, 15 different game studios across all categories of gaming that we ended up owning so that we can not only understand what the game developer needs from marketing platform and a monetization platform, but also use their data inside our model. And so, we we bought the studios, we launched our first Axon 1 model. That Axon name is the name of our our advertising model. The first one was more traditional machine learning, but huge upgrade over the first system. And revenue growth started happening really quickly once you brought that to market, but we couldn't do it without owning these studios ourselves and being able to feed data into the model to train it. We moved really quick. We have an idea and we just go. And so one of the nice things at KKR was they were actually supportive of this. It wasn't like let's start analyzing like what it is we're doing. It's like Adam's got an idea, let let's go behind him and we just made this investment. So we got to believe he knows what he's doing. And so it was idea to first acquisition almost right away. It was we have a studio to go buy. Um it was a studio called PeopleFun. And then we were building up our own studio internally, too. And then everyone was supportive cuz the logic was really, really sound. Now the the fear and why a bigger board might have said, "Hey, we don't want to do this." is we're platform servicing gaming companies. And everything I've done in my career is be transparent all the time. Like your reputation and trust can break down at any moment if you violate trust on the other side of things. And so I'm a very transparent person and we end up in this space where we're servicing the gaming clients and we're now going we're going to compete with you. And so in their mind it was okay, the ad platform's becoming now vertically integrated competitor. And the proof for this was Amazon does this all the time, right? Like why wouldn't the distribution platform build their own brands? In our case it was build the brands to get data. It's very hard if you're a client on the other side and you see this to believe that they're still they've still got your best interests in mind. And so that would have been the reason you wouldn't choose to do something like this. That was what ended up over time. So like the the fact There there was a moment in time where a lot of people stopped trusting us in the in the space because our games were growing really quickly. We were feeding their data into our model. And so, the early adopters of our apps on one platform were our own games. So, then you start seeing our games grow. On the one hand, you had companies that go, "You're competing with us." who just chose not to work with us. We said, "Fine. That that's not only logical, it's okay cuz we made this bet and we're going to live with it and it's really good for our business potentially." On the other hand, you had smaller businesses small to medium-sized inside gaming. They just need any win they can get. And so, we ended up with these companies that then said, "Okay, your own studios are sharing your model data. We're going to share your model data and help us grow our business." Yes and no, right? Like, you go, "That's great." But then if I feed you data, that's helping your own game, so am I fueling a competitor to become better? That's what the concern would be, right? It's like, you're using my data to effectively get into the space and compete with me. Now, you have to know who's on the other side of it. One other interesting data point about me is, growing up I played games, but as an adult I hate games. I don't I don't spend a minute playing games. So, like, it wasn't like I I mean, it just consumes time. It's addictive. Like, it it's like, for me, myself, like, I don't have the patience to like get good at games. So, therefore, I suck at games and I don't like sucking at things, so I don't play games. You talk to these game developers and they they like what they had to really realize is like, the company run through me can't be a game developer. I don't even care about the game content. I cared about the data and I cared about the distribution platform. So, once they started realizing, "Okay, they're never going to be a good game developer. This is clearly a strategy. It's a means to an end." it started changing things. But it required I I know a lot of the game developers in the space. A lot of us are now quite close and over time, like, over the last decade, a lot of us have grown up in this industry together, so we've become close. So, it just required a dinner, a lunch, sit down, ask me whatever question you want. I'll say in hindsight, the mistake I made is I'm very transparent when I sit down with people and they start asking me questions, but I didn't convey to the community of game developers directly and explicitly, we're getting into gaming business and here's why. And so, people started finding out on their own and then thinking there was some malicious intent when there never was. But then once they sat down with me, it was like, okay, now we can overcome that and like get down to business. How are we going to grow your business? And so, as a company, we've always had a belief that if our clients are improving their business, we're whole. We're we're good to go. I mean, the gaming in the mobile gaming space, there's very little VC funding. A lot of these companies are bootstrap businesses. The founders are the key people. The founders are the marketers. Um most of the companies we work with are not even in the states. They're in Europe and Asia. And so, these types of businesses, they need partners that are there to help them. And for a moment's time, because I was not proactive in in talking about what is our strategy, I broke trust down. Then we got back together, sat down with the the companies in the space. And trust comes back really quickly when you you realize like it's a communication error, not a desire to deceive. We ended up going forward and all of these companies over from when we launched Auction One Model in 2020 to today, they all work with us in much, much bigger scale. And our platform's become the dominant one in gaming. We're we're probably I I don't I don't know how big, but maybe even almost the over 50% of the marketing dollars in the space back to our platform because the the developers on the other side are seeing the best best best returns that they've ever seen in their existence on the dollars that they're spending on our platform. So, their businesses are growing really healthy rates. It's been powered by this desire to by from us to build this technology the way we did and and interest from them to trust us as the platform to do right by them. And that evolved over the last few years, but it's gotten to a really great place. Align with the fact that we were never going to be good at gaming. We got to a place where the business had grown so much that every advertiser in mobile gaming was now starting to connect into our models and and work with us the right way. Once we got to that point, we realized we don't need these studios. Headache, it's a distraction. As you said, the the ad core team on advertising is only 400 people. Well, to run studios is a huge lift. You need a lot of head count. There were 1,500 plus people on those businesses and it was distributed over the world. So, we got to a place where we said we just can't manage this. It's not in our DNA. We don't know what's going on. We're not paying attention to it. And we ended up um selling it off to this company whose whose founder I've known a very long time. It's a great guy. This company triple.dot out of the UK. They bought the whole thing and that was important to us. We didn't want to go through and say we we own 15 studios. We're going to sell off one at a time just to make a profit. So, instead we just priced it down enough to make it enticing to a company like triple.dot to just take the whole thing off our hands. And we did that. I believe that was about a year ago. From beginning to end, I want to say five years roughly. You have to have one of the craziest like employee to market cap ratios. So, how are you able to run a $150 billion company that's making what, a couple billion a year in cash flow? I mean, run Okay, so 5 billion or there a little bit more uh per year with 400 350 to 400 employees in the main business. How are you able to do that? Why'd you tank in 2022? Okay. Did you go public So, wait, go into more of the thinking cuz this sounds like it's going over several years from 22 to 23 to 24. We're only talking about the ads business. We're not talking about the other uh you own like a SaaS company that I think has Let's Let's just focus on this. So, now you're you know, 350 to 400 of them today. What were you back then? That's excellent. Do you think there's a certain size, like a limit to how many If you want to If you're running a company, you want a team of all A players. Do you think there's a an upper bound to how many employees you can have? So this leads me to another thing that I find remarkable about you that even to this day, if anybody wants to hire anybody in the company, they have to get an approval directly from you. I really think what's remarkable to you is you just took this insight that everybody agrees on, right? That the value of your company is just the talent, the level of talent that you're you have in your employees, your team. You only want to work with A players. There is some kind of upper bound to it. I remember the reason I asked you the question earlier is because like Steve Jobs said that he had only ever seen one company in his entire history and his entire experience that was all full of A players. He says it was Pixar. He said but they had 400 employees and at the time Apple had like 3,000. So he says of course I have 3,000 employees, they can't all be be A players. But what's remarkable is everybody agrees on it but no one actually does the the difficult job of like making sure that you don't alienate these A players and and essentially surrounding them with Bs and Cs and other and other less talented people because you jump to the right conclusion. They're going to leave and in many cases they'll just go start their own company. So, we've now spent many hours talking about the way you think about building your business. The way I would describe it, if I had to like distill it down to just handful of words, I would say it's ruthless efficiency and like hyper competence. Is this in response, cuz it's not common, especially in the startup VC world where it's just like they just throw money at everything, they have thousands of people or over hire and everything else. Is is your like ruthless efficiency and hyper competence in response to something or is this just how you've always been? I'm curious what happened to your better funded competitors that started either around the same time or after yours? The reason I ask is I talked to Michael Dell about this. You know, there's a million people doing what he was trying to do, and in many cases they were doing it in the same city that he was trying to do. All better funded than him. Some had raised, you know, I think over a hundred million dollars. He smoked every single one of them. What happened to these competitors you mentioned earlier? You know with your if your Axon model is good or not immediately, right? Cuz you're like essentially the in you're in the business of selling revenue, I think is the way you put it, right? So, explain when you knew it was working the way it is now and what was happening before. Like what was the the model before that? I remember seeing a graph this like inflection point. When somebody might have heard you say, "Hey, you know, we we're not we don't we don't have like specific market cap goals, but we do have a plan in place for some of the key players for compensation that goes all the way up to like a trillion-dollar market cap." Might sound a little crazy, but not if it seems like you just have a ton of tailwinds, like AI tailwinds for the specific business, not only the actual part what the product is, but your like ruthless efficiency and hyper competence with only like having a small team of A players. Obviously, you have somebody who really gets it, they could be it's not one employee, you don't really have 400 employees, you could have 40, you know, them doing the work of 4,000. Explain like how just even the last like three or four years, how AI has transformed the way you're running the business and the impact on the business. So, you go from gaming to e-commerce. You literally just My next question was going to be like, "Well, what other kind of businesses do you want to expand out to?" You really believe that you can do this for every single segment? When you have this many people who are playing games and they're middle-aged, the only limit is the technology for targeting and the customer on the other side. And I fundamentally believe the biggest opportunity we have in as a business is help the small to medium-sized businesses. We don't have a sales force, so eventually we'll get the large enterprises cuz they're going to come to a platform as big as ours, but we're not out there selling them. We want to help the small to medium-sized businesses. What made us really successful in gaming was going to the companies that really didn't have much support that most of the other businesses are getting, like they're 10 people. Let's work together, let us grow your business. There's a company that recently exited out of Turkey in mobile gaming that I believe was like 10 15 people and sold for a billion dollars. They sold under 1 year from launch for a billion dollars. Almost all of their business growth on the marketing and monetization side was built on our platform and they had a fantastic game that was super innovative around that, but because our platform's built for this, we were able to help a company like that and that makes us feel really good. So, if we can translate that into these other categories, the local laundromat or to the company that is spins up a Shopify store and is selling lipstick and has some cool spin on selling lipstick and they launch on our platform, they can really ramp up, we'll feel really good about ourselves. Eventually we hope to tackle the consumer market and enterprise market. There's no reason not to cuz again on the other side of your billion users and they're adults, they have jobs as well. If we get to the point where our technology can do the targeting across every consumer category, I'm pretty sure it's going to have no limit when it comes to the enterprise side as well. Yeah, thanks for having me. This was really cool.

FAQ

A. 英語原文・自動字幕について
なぜ英語が「自動字幕ベース」なの? 公式の文字起こしは?

このページの英語は、YouTube が生成した automatic captions(ASR=音声認識) を yt-dlp で取得したものです。人手で整えた公式トランスクリプトではありません。 公式サイト(davidsenra.com)にはエピソード要約・ショーノートはありますが、取得時点では全文の公式文字起こしは前提にしていません。 参照: https://www.youtube.com/watch?v=SCBFXL_Sn44 / https://www.davidsenra.com

自動字幕の精度はどれくらい? どこが壊れやすい?

一般に YouTube 自動字幕の精度は条件で大きく変わります。 - 条件が良いと 85–95% 前後まで上がる例もある - アクセント・早口・専門用語・固有名詞・複数話者 で一気に落ちる このエピソードはインタビュー(2人会話)で比較的クリアですが、金融・広告・規制・人名が多く、名前の誤変換が起きやすいタイプです。

この記事の英語で実際に見つかった/起きやすい誤変換は?

自動字幕原文に出やすい典型例です(日本語本文側では多くを修正済み)。

| 字幕に出た/出やすい表記 | おそらく正しい表記 | |---|---| | Apploven / App Lovin | AppLovin | | Axon / Auction One | Axon(広告モデル。Axon 1 / Axon 2) | | Orient Hontai / Orient Hongtai 等 | Orient Hontai Capital(中国PE。字幕ゆれ) | | CFIUS / CFIAS | CFIUS(Committee on Foreign Investment in the United States) | | Harold Finn / Harold Sham | Harold Finn(KKR。字幕ゆれあり) | | Basel / Basil Shikin | Basil / Basel Shikin(CTO。本編で両方に聞こえる) | | People Fun | PeopleFun(買収スタジオ) | | triple.dot / Triple Dot | triple.dot(スタジオ売却先) | | Raph / Raft | Raph(長期共同創業級の仲間) | | Eduardo Vivas | Eduardo Vivas(Raphの兄、初期ボード) | | Giovanni | Giovanni(後にCTO、Axon 2) | | AdMob | AdMob(Googleが買収したモバイル広告) | | sendora / enra | senra(スポンサースラッグの崩れ) | | Deal | Deel(スポンサー) | | RAMP | Ramp(スポンサー) |

つまり英語欄は「音声の生ログ」に近く、学習用の対照テキストとしては有用でも、引用の一次ソースとしては動画そのものを優先してください。

日本語と英語は1対1で完全対応している?

完全対応ではありません。 - 日本語は読みやすさのため発言をまとめ・整文している - 英語は自動字幕の切れ目が多く、発言数が日本語より細かい場合がある - ページでは話者(センラ/フォロウギ)を手がかりに、順番で英語を紐づけている そのため「この日本語のこの1文 = この英語のこの1文」とは限らず、だいたい同じ話題の塊として読むのが正しい使い方です。

上部トグル(英語原文 / 少し大きく)は何のため?

- 英語原文 ON/OFF … まず日本語だけ読みたい人向け。英語は約9.5pxの補助情報 - 英語を少し大きく … 約11px前後。対照しながら精読したいとき用 英語をメインにすると誤変換に引きずられやすいので、デフォルトは「日本語メイン・英語は極小」にしています。

スポンサー部分は?

本文から省略しています。音声上は Ramp / Deel / Axon(AppLovin) 等の読み上げが入っています。番組エンディングの Founders 案内も本文では省略しています。

B. 人物・会社・固有名詞
Adam Foroughi / AppLovin とは?

Adam Foroughi(アダム・フォロウギ)AppLovin の創業者・CEO。広告・ゲーム領域で3社目の事業として、2011〜2012年ごろモバイル向けに創業。最初の1年半はブートストラップ、2012年3月にアドネットワークとしてローンチし、同年11月にエンジェル規模の調達。本編では取締役会なしで約6年走り、2018年にKKR関連の再資本化で初めてボードを置いた、と語る。 AppLovin はモバイル広告プラットフォーム(製品ブランド Axon)を主軸とし、一時は多数のゲームスタジオも保有した(後に売却)。本編の時価総額・キャッシュフロー数字は話者の口頭ベースで、IRと完全一致する保証はありません。

David Senra(センラ)とは?

起業家伝記を扱う Founders ポッドキャストで知られるホスト。本番組 David Senra Show では創業者インタビューを行う。本編では買い戻し・資本市場・Aプレイヤー文化・Axonのパフォーマンスマーケを深掘りしている。

自社株買い(買い戻し)の話の要点は?

2021年4月IPO後、株が約92%下落(本編: 115ドル→9ドル、時価約38億ドルの底)する一方、EBITDAは2021年約7億→2022年10億超と伸びた、と語られる。 市場が株を嫌うなら自社で買う、という方針で、公開市場ランダム買いではなく、将来売るPE等と合意して時間をかけて買い戻した。約18ヶ月で約60億ドル規模(自社キャッシュ+レバレッジ)。その結果として約500〜600億ドル相当の価値創出になった、というのが本編の主張。数字は口頭ベース。

VCに断られた「100万ドル/400万バリュエーション」とは?

2012年春ごろ、会社の約25%を100万ドルで渡す提案をトップVCに全部断られた、という回想。広告+ゲーム+Google/Facebook/Amazonとの競争懸念が理由と本人は整理。同じ頃・直後に同アイデアの競合にはVCマネーが入り、動機になった、と語る。

前の2つの広告事業とDTC失敗は?

2005年・2008年ごろ、デスクトップ/ソーシャル広告で成功し売却。モバイルへのトラフィック移行をデータで先に見えた。最初は広告を避け、2011年に出会い系・ファッションのDTCアプリを出して失敗。3つ目の「アプリ発見アプリ」の推薦の反応率が高く、SDK経由の広告ネットワーク(2012年3月)にピボットした、という創業ストーリー。

AdMobとの違いは?

Googleが2008年に買収した AdMob が当時の代替。ブランド向けが中心になりがちだったのに対し、AppLovinは開発者向け——配信面としての掲載+ユーザー獲得としてのパフォーマンス広告——を両側に置き、ブランド接待型ではなく「営業なし・高価値プロダクト」を志向した、と説明される。

パフォーマンスマーケ/「広告主をアービトラージャーに」とは?

1ドル使うと、ビジネスモデルの回収期間の範囲でそれ以上の売上(または明確なリターン)が見える状態にすること。そうなればスケールの制約はほぼ銀行残高だけになる、という定義。Facebookの一部の強みに近い、と本人は言う。Axon 2(ディープラーニング)以降、プラグインしてすぐその状態に近づけるようになった、が2023年4月ごろの変曲点の説明。

Axon 1 / Axon 2 とは?

AppLovinの広告予測・配信モデルの世代名。 - 初期: ルールベース(例: このゲームをする人に別ゲームを推す) - Axon 1: 従来型機械学習(スタジオデータ取得後) - Axon 2: ディープラーニング。本編では2023年4月ごろのロールアウトが事業・株価の大きな転換点とされる 「広告をコンテンツにする」=ユーザーにとって本当に欲しいものの発見になるほど見せられる、という文脈で語られる。

ゲームスタジオ買収と triple.dot 売却は?

広告主データが得られなかったため、2018年以降 PeopleFun など14〜15スタジオを買い、垂直統合でデータをモデルに入れた。自社ゲームが伸びると「競合になる」と離れる顧客もいたが、透明な説明と結果で信頼を回復。データ目的が達成され、広告コア(約400人)に対してスタジオ人員(1500人超)が散漫になり、triple.dot にまとめて売却(所有約5年、本編時点で約1年前)。ゲームコンテンツ自体への興味は薄い、と本人が明言。

中国PE・CFIUS・コンバーチブルノートの話は?

2016年ごろ Orient Hontai Capital と、約14億バリュエーションに対する約10億ドル投資(当初は支配権に近い構造・中国上場想定)を発表。CFIUS 等のクロスボーダー審査が長く続き、地政学・データ・国家安全保障懸念で頓挫しそうになった。ボードなしで進めたことを後悔しつつ、投資側への忠誠から、70%支配のエクイティではなく、10%相当のコンバーチブル/非支配構造に組み替え、株主に約10億を配当した、と語る。その後 KKR 等の米国資本とデットで中国側コンバーチブルを整理し、2018年に初ボードを設置。

KKR / Harold Finn / 初ボードとは?

2018年夏ごろの再資本化で登場。Harold Finn とチームが「変な社名」でも中身を見て乗った、という逸話。ボードは Adam、Eduardo Vivas(Raphの兄)、Harold の3人から。IPOは2021年4月、バリュエーション約200億・調達約20億・フロート約7〜8%(低すぎてボラ要因だった、と本人)。

Raph とは?

高校中退、16歳から Adam の前職から約18年働くキーパーソン。BD・ディール・記憶力(会話の逐語記憶)が強み。最初の6ヶ月はPalo Altoのオフィスに二段ベッドで住んだ。採用哲学「肩にチップがある人」の象徴として語られる。いまもマーケ/グロース側でハングリー、という描写。

「肩にチップ」とイランからの移住は?

家族がイランから追われ、4歳で米国へ。父はイランで大規模不動産デベロッパーとして成功していたが、移住後「一部が失われた」ように見えた、という幼少期の動機。成功の到達点は動くターゲットで、常に押せ、という自己像。採用でも同様の動機がある人を好んだ、と語る。

従業員350〜400人でFCF約50億/時価千億級、とは?

本編口頭: コア広告事業で約350〜400人、四半期キャッシュ約13億→年50億超、時価はピーク約2500億・本編時点で約1400億などと語られる。数字は口頭・時点依存で、最新IRを優先すること。 仕組みの話としては、(1) 2022年の株価崩壊後、エクイティをキー約100人超に集中、(2) 2023〜24に事業成長中でも約40%解雇してAに蒸留、(3) CRO/COOなしの薄いCスイート、(4) 採用はCEO承認でバックフィル自動停止、(5) AI/LLMでコード生成比率が高い、などが本編の説明軸。

Aプレイヤー文化と「全採用はCEO承認」は?

Jobsが「Aだけの会社はPixar(当時約400人)くらいしか見たことがない」と言った話と重ね、B・Cで囲むとAが辞める、という実行論。Giovanni(後CTO)の「なぜこの人/このプロセス?」という詰問が、プロセス肥大の見直しとカットのきっかけになった、と語られる。辞めた席の自動バックフィルを止め、本当に要る採用だけCEOが通す。

共同創業者の入れ替わり・「常に最良の人を席に」は?

当初CTOの John がチームを作り、Basel が交代。本編直前も Basel が別役割へ、その下がCTOへ、と同パターン。安住すると下の優秀層が辞めて起業する——会社と個人の両方への義務として席を入れ替える、という方針。公開企業化で要る人材タイプも変わった、と説明。

eコマース拡張とSMBは?

ゲーム広告から、同じ10億超オーディエンスにeコマース等のプロダクト広告を載せる拡張。セールスフォースを持たず、大企業より中小・ indies(例: 10〜15人規模のゲーム会社の急成長)を助けるのがDNA、と語る。最終的にはB2B/エンタープライズも技術的に限界はない、という展望。

本編の数字をそのまま投資判断に使っていい?

いいえ。時価総額・株価・EBITDA・買い戻し額・所有比率・人員・コードのLLM比率などはインタビュー時点の口頭です。正確な数値は AppLovin の公式IR・SEC提出書類・決算資料を当たってください。

C. 関連ファイル

- 対話本文: article-ja.md - 見た目: article-ja.html - 日英対照: article-bilingual.md - FAQ: FAQ-ja.md - メタ: meta.json - 英語整形: transcript_en_clean.txt