Aug 14, 2026 · 1h 26m · sourcery
Inside AppLovin’s $100B Ad Engine
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
AppLovin CEO Adam Foroughi and CTO Gio Ge join host Molly O'Shea on Sourcery to discuss the engineering architecture, lean organizational culture, and performance advertising strategy driving their rebound from a 92% market drop toward a potential trillion-dollar valuation.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Molly holds 16.3% of the talking time here. How this is scored →
Molly as informed peer 4.2 Guest teaching 3.3 Guest disagreement 2.0 Molly pushing back 2.1
speaking balance: gold is Molly, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 48:28 Gio rejects the AI rebuild premise
Gio directly dismisses the host's premise that companies must rebuild their architectures for the AI era, stating he has no desire to rebuild and that AI only layers on existing foundations.
Hardest push from Molly ▶ 1:11:02 Molly challenges Adam on feeling no excitementWhen Adam claims he simply doesn't feel excitement about milestones or market success, Molly immediately challenges his framing, asking incredulously what he feels and whether he feels anything at all.
Biggest teaching moment ▶ 17:12 Gio corrects Molly on token maxingWhen Molly tries to use LLM jargon ('So you're not token maxing'), Gio immediately cuts in and educates her that token maxing belongs to large language models rather than matrix recommendation systems.
Molly holds their own ▶ 1:24:47 Molly names gaming-to-defense founder precedentsMolly demonstrates deep ecosystem knowledge by connecting gaming engineering origins to defense tech transitions, citing Torsten Reil of Helsing and Palmer Luckey of Anduril to press Adam on his future.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Molly as informed peer | Guest teaching | Guest disagreement | Molly pushing back | Why |
|---|---|---|---|---|---|---|
| Gio's Arrival and AppLovin's Lean Engineering Culture | 4 | 3 | 2 | 2 | Molly opens by framing AppLovin's turnaround around Gio joining as an inquisitor asking hard questions. Gio slightly reframes that he didn't start by asking questions but by quietly writing code and observing that junior engineers lacked business context. Both Adam and Gio explain the lean, IC-heavy culture in collaborative detail. | |
| Rebuilding the Algorithmic Engine and Axon 2's Origins | 3 | 2 | 1 | 1 | Molly asks direct setup questions about team size and the technical architecture sprint for Axon 2. Gio recounts his 12-hour offline coding flight to Italy and work ethic shift, with Adam reinforcing the rapid iteration culture. The exchange is warm and storytelling-driven. | |
| Technical Architecture and Mathematical Mechanics of Axon 2 | 3 | 7 | 3 | 2 | Gio breaks down Axon 2's mathematical shift from tree-based clustering to semantic embeddings and deep neural networks optimized for GPU GEMM operations. When Molly tries to interject with buzzword slang ('So you're not token maxing'), Gio explicitly corrects her that token maxing applies strictly to LLMs, prompting Adam to joke about needing translation. | |
| Performance Business Model and Expansion into E-Commerce | 4 | 3 | 1 | 2 | Molly inquires about conversion funnels, data capture during 60-second interactive ads, and expanding into e-commerce. Adam explains closed-loop ROAS dynamics and contrasts generalist embeddings with verticalized assumptions, citing how casual mobile gamers fit prime buying demographics. | |
| Napkin Prototyping and Engineering Without Product Managers | 3 | 2 | 1 | 1 | Molly asks how testing and prototyping work in practice. Gio and Adam share an anecdote about prototyping their e-commerce data pipeline on a breakfast napkin in Vegas, underscoring their deliberate choice to operate without product managers. | |
| Sponsor Segment: Brex Corporate Finance Platform | 4 | 4 | 2 | 2 | Following an ad break, Molly asks how engineering interfaces with advertisers without intermediary account layers. Gio and Adam clarify that while business teams maintain client relationships, engineers operate with hard ROI formulas and clear context rather than micromanagement. | |
| Humans on Top of AI and Engineering Accountability | 5 | 3 | 2 | 2 | Molly cites her conversation with Max Levchin regarding whether engineers still need to learn coding. Gio insists they do because human engineers sit 'on top of AI' and must be held accountable for system decisions rather than blaming AI, while Adam discusses how engineers must generate algorithmic alpha rather than out-of-the-box prompting. | |
| Creative Strategy, User Attention, and AI Video Limits | 4 | 4 | 2 | 2 | Molly asks about creative strategy, referencing agency claims about producing hundreds of videos and personal anecdotes around viral UGC. Adam corrects the social UGC premise by emphasizing that AppLovin's long-form, 60-second interactive playables cater to older, non-TikTok demographics requiring sustained brand trust and varied creative concepts. | |
| Building for the AI Era and Avoiding Low-Taste Slop | 4 | 4 | 3 | 2 | Molly asks how companies should rebuild for the AI era. Gio pushes back against the premise, stating AppLovin has no need to rebuild what they already made, but rather builds on top of it. Adam notes that 99% of people use AI as a basic search engine and warns against spending budgets to produce low-taste internal slop. | |
| Sponsor Segment: VCX by Fundrise Venture Ticker | 4 | 3 | 2 | 2 | Following mid-roll sponsor reads and banter about Claude, Molly asks whether AppLovin plans to expand advertising into LLM chatbots. Adam distinguishes bottom-of-funnel search intent from AppLovin's top-of-funnel discovery engine, indicating connected TV and open web video are far better fits than conversational bots. | |
| The Math of a Trillion-Dollar Valuation and Market Turnaround | 5 | 3 | 2 | 3 | Molly poses a direct question on how AppLovin intends to hit a trillion-dollar valuation after dropping over 90% post-IPO. Adam provides cash flow arithmetic ($30B+ annual free cash flow requirement) and explains their capital allocation playbook of cutting IR, avoiding excessive stock dilution, and repurchasing undervalued shares. | |
| Operational Mindset, Execution Discipline, and the Singularity | 4 | 3 | 3 | 4 | When Molly asks what excites them, Adam deadpans that he does not experience excitement or emotional highs. When Molly asks about the AI singularity, both guests express skepticism toward Twitter hype, with Gio rejecting defeatism to defend human agency and Adam noting real consumers enjoy active shopping experiences. | |
| Hottest Takes on Engineering Taste and AI Economics | 6 | 3 | 2 | 2 | Molly prompts both guests for hot takes and connects Gio's point on taste to Mickey Malka's critique of Silicon Valley losing aesthetic sensibility. Gio emphasizes that engineering taste means knowing what not to build rather than blindly copying giants, while Adam argues AI faces public backlash because marketing fails to show tangible economic upside. | |
| Personal Performance, Leadership Traits, and AppLovin's Future | 6 | 2 | 2 | 2 | Molly asks about Michael Barton's description of Adam being completely locked in, and later cites peer founders moving from gaming to defense (Torsten/Helsing and Palmer Luckey/Anduril) to ask what comes next for Adam. Adam laughs at her native Brex ad integration and affirms his commitment to leading AppLovin long-term. |
Statements from this episode (25)
AppLovin CEO: Gio Ge built Axon 2 to replace outdated tech
“We had a model that powered our advertising business axon one model, and it was written on outdated machine learning techniques. And so we were looking for a researcher to come in and help us architect the axon two model, which we released. I don't know, proba…”
Foroughi: Unprofitable Advertising Companies Should Not Be in Business
“An advertising company, if you can't make money, then you probably shouldn't be in business. Doesn't make any sense, because you're trying to deliver value to advertisers, you get paid for the value you deliver, and in theory, that should generate you somethin…”
Former AppLovin CTO Basil personally wrote roughly 60% of the company's code
“Basel probably wrote, I don't know, 60% of the code of the company like himself.”
AppLovin cut its headcount from 600 to 400 and kept it flat
“I'd say the company's been around of 400 people for a while. And About two to three years ago, we probably cut it down from 600 to 400, and we haven't really changed all that much since, and people come and go, but that's what you got.”
Ge: Neural network ad models run far more efficiently on modern GPUs than decision trees
“The neural network, the benefit of neural network is, it's made up of a lot of standard GMM, general matrix multiplication, which, you know, is a famous, the GPU. GPU is highly optimized for this kind of operation. But the selection tree, you know, the old gen…”
Foroughi: AppLovin users spend roughly 60 seconds actively playing mini-game ads
“The ads inside games are much more similar to what television is, but specifically for games, You have somewhere around 60 seconds that people sit on an ad, and they're not just sitting there doing nothing, they're playing mini games that we serve them. So the…”
Foroughi: Heavy casual mobile gamers skew female, aged 30 to 50
“And the people who were playing mobile casual games turned out to be a lot of heads of households. It skews slightly female, but it is this sort of 30 to 50 year old age group that's the heavy mobile casual gamer.”
AppLovin operates with almost no product managers, relying entirely on engineers
“One of the cultural philosophies we had is almost no product management and have engineers that are competent enough about business problems to just write their own architecture, write their own product to solve business problems that maybe not even a business…”
Foroughi: AppLovin engineers and interns can push code in first week
“So one of the things that's a little unique is like whether geo is an intern or a full-time employee just started maybe even right out of school they can push code within the first week of being here.”
Foroughi: AppLovin has no training manuals or formal onboarding
“We don't have the traditional training processes that bring people along. There's no training manual. There's no come in and, like, someone's gonna, like, tell you what to do.”
AppLovin strictly holds engineers accountable for any AI-generated code decisions
“You can use AI however you want.
But in the end of the day, you as a human is held accountable for every decision that your AI made for you.
So it is absolutely not acceptable when I come to an engineer and say, hey, why you did, why you implemented this syste…”
Adam Foroughi: Generic AI usage eliminates alpha and commoditizes products
“Like if the AI starts writing all the ad creative, then all of it's going to eventually look the same and the user won't respond to the ad.
So how do you create alpha?
Well, you have to have really smart people that understand systems, understand what they're …”
Foroughi: Creative investment creates alpha when ad systems are fully automated
“The advertisers that invest in creative and are good at it, whether they're using AI tools to develop them or they're using people to develop them, can create lifts in an advertising system where everything else is automated.”
Foroughi: AI cannot yet autonomously generate high-performing long video ads
“It's not at a place where you're just going to get a massive amount of inflow of advertisements that are diverse and creatively inspired and going to create lifts and campaigns, but their tools are available now so that humans can create ads much faster and mu…”
Gio: AppLovin is developing fully automated AI ad creative generation
“We have to create a tool that can fully automate everything. The other choice is you just create a tool that allows advertisers to interact with, to allow them to collaborate with the AI tool to create their own content. So right now we see that the second typ…”
Gio: Wasteful ideas and bad taste offset AI's productivity gains
“I think a big, big mistake that people are making is that because of the AI, building things become easier. People are making a lot of bad decisions, bad tastes, wasteful ideas that actually offset the empowerment of AI.”
Foroughi: Non-AI-native legacy companies must replace nearly everyone to successfully adapt
“Technically you almost have to replace near everyone and rebuild the culture up to be able to deal with the world as it is today, but that's not really possible for most companies. And so there's no clear answer for how those companies can get there.”
Foroughi: Frontier AI labs will not become dominant operating systems
“And if we ever got into advertising with some of the chatbots or large language models, you'd say, would one of these frontier labs become an operating system? It doesn't look like they're going that way. So like, even there, you're going to have fragmentation…”
AppLovin generates over $7B EBITDA run rate with 75% cash conversion
“We probably on an EBITDA basis, I think we're over a seven billion dollar run rate, and we generate somewhere around 75% cash off the EBITDA dollar.”
Foroughi: Mobile gaming ad spend cannot support a $1T valuation
“The mobile gaming category is not big enough in terms of user acquisition dollars to support the type of cash flow we need to be valued at a trillion dollars.”
AppLovin dropped 92% post-IPO before rebounding to a $100B valuation
“We dropped 92% in the first 18 months of being public. That, that dream seemed very, very far away, but post-Axon, too, we recovered, and now we're a hundred-billion-dollar company.”
AppLovin halted investor relations for over a year during its market bottom
“We actually didn't do any investor relations for well over a year at the bottom, because what were we going to convey to investors is we're rebuilding our technology stock is dirt cheap, but nobody buys something that's dirt cheap.”
Foroughi: AI chatbots will not eliminate ad-driven e-commerce
“We get questions like, well, isn't the chatbot going to tell everyone what to buy and just complete the transaction and no one will ever do e-commerce off of an ad again. And what that leaves a sight of is that people actually like to shop and people want to l…”
Foroughi: Public will not support AI without use cases beyond search
“Until people start understanding what the average person can have as positive use cases that are incremental to search, they're not going to be proponents of AI, and there's no reason to expect them to be.”
Foroughi: AppLovin CEO avoids angel investing, conferences, and social media
“I'm not out there super active on social media. I'm not out there making a whole bunch of other investments. I'm not out there speaking at a lot of different conferences. I'm basically a hundred percent wired in to try to do the best job that I can with this.”