newsfilter.io
Interview, Fireside Chat

Inside AppLovin’s $100B Ad Engine

  • Growth Trajectory & Valuation:

    • AppLovin dropped 92% in the first 18 months post-IPO but recovered post-Axon 2 release.
    • The company reached a $100 billion market capitalization.
    • Current run-rate EBITDA exceeds $7 billion, generating approximately $5.25 billion in cash flow annually.
    • To achieve a $1 trillion valuation, the company targets $30 billion+ in annual cash flow.
    • Adam Mosseri (CEO) notes the company has "dropped 92% in the first 18 months of being public" but recovered.
  • Technical Architecture (Axon 2):

    • Gio (CTO) and Basil (former CTO) rebuilt the core recommendation engine from scratch, abandoning the outdated Axon 1 model.
    • The new model replaces tree-based algorithms with "semantic embedding" to capture dynamic user-item interactions.
    • The Axon 2 architecture utilizes GPU-optimized GEMM operations, making it cheaper to run than the previous selection tree model.
    • The team operates with ~100 engineers, maintaining the same headcount as three years prior despite significant business scaling.
    • Engineers are required to write code without AI assistance in early onboarding to ensure they can audit and correct AI-generated errors.
  • Strategic Shifts & Expansion:

    • AppLovin expanded from mobile gaming advertising to e-commerce and general consumer verticals approximately one year after Axon 2's release.
    • The company acquired "pixel" data from advertiser websites to improve user profiling beyond just gaming behavior.
    • Business expansion was driven by engineering prototyping rather than traditional product management specs; a napkin sketch at a Google conference led to the first e-commerce data engine.
    • The company deliberately chose not to build a large sales force, relying instead on a self-serve model where advertisers pay based on proven ROI.
    • Advertisers are encouraged to create 20 to 50+ ad variations per campaign to find high-performing creative concepts.
    • AppLovin does not currently plan to build fully automated AI ad creation tools due to the high risk of brand trust erosion; they prefer AI-assisted human workflows.
  • Operational Philosophy & Culture:

    • Gio attributes Axon 2's success largely to "what we decided not to do" rather than what they did, focusing on avoiding resource-intensive replication of giant competitors.
    • The company avoids "rebuilding" for the sake of AI; instead, they use AI to scale output with their existing lean team.
    • Onboarding involves pushing code to production within the first week to provide immediate feedback loops and dopamine hits.
    • The culture emphasizes "taste" (deciding what not to build) over raw technical capability as the primary differentiator.
    • The company buys back its own shares aggressively, viewing itself as its best investor during market downturns.
  • AI Integration & Future Outlook:

    • Gio's philosophy: "I don't want our engineers to sit next to AI. I want our engineers to sit on top of AI."
    • Engineers must retain coding literacy to judge AI decisions; AI is not accepted as a final decision-maker for core infrastructure.
    • AppLovin predicts AI will not dominate bottom-of-funnel search ads but will excel in top-of-funnel discovery scenarios (e.g., Connected TV, open web video).
    • The company plans to scale operations without proportional headcount increases by automating business processes and using AI for data analysis.
    • CEO Adam believes the "singularity" is likely a Twitter myth and that human interaction remains essential for long-term business success.
    • Future growth relies on moving beyond the mobile gaming TAM to adjacent consumer verticals and eventually B2B/enterprise solutions.
  • Leadership & Talent Management:

    • Gio joined in November 2022 when the stock dropped 30%; he viewed the dip as an equity opportunity rather than a risk.
    • Adam Mosseri prioritizes hiring for "intelligence" and "low ego," specifically looking for individuals who question themselves and handle adversity well.
    • The leadership team believes "taste" is an underestimated skill in the AI era, distinguishing successful companies from those that produce "slop."
    • Adam optimizes his life to remove distractions, focusing 100% on building the business and avoiding social media/conference circuits.
    • Gio cites a "desire not to disappoint" as his primary motivational driver during his early career.
  • Market Position & Risks:

    • AppLovin faces a duopoly in mobile (Apple/Google) but targets fragmentation in Connected TV and the open web for expansion.
    • The company warns that many public SaaS companies will fail to recover from 90%+ stock drops due to an inability to fund their own turnaround or scale revenue without massive sales headcount increases.
    • Gio notes that companies with high tech debt and resistance to rapid iteration risk obsolescence in the AI era.
    • The company expects to maintain a lean structure, avoiding the "bloat" typical of companies scaling alongside revenue growth.