Investors should prioritize discovery-based advertising platforms over traditional search, as browsing-driven consumer platforms are far more insulated from generative AI disruption. AppLovin (APP) represents a high-conviction growth opportunity as its deep-learning recommendation engine expands from mobile gaming into e-commerce advertising, backed by industry-leading 84% to 85% EBITDA margins and strong cash generation. Meta Platforms (META) offers similar structural resilience, using entertainment-driven social media to capture high-margin impulse shopping that AI chatbots cannot easily replicate. In contrast, investors should exercise caution with Alphabet (GOOGL), as emerging conversational AI models directly threaten its core intent-based search advertising franchise.

AppLovin (APP)

• AppLovin operates an algorithmic advertising platform powering monetization across more than 100,000 mobile games, reaching over 1 billion daily active adult users.
  • The company operates within an estimated $50 billion annual mobile gaming ad ecosystem, with its own platform handling approximately $20 billion in ad spend following ~60% year-over-year growth.
  • The core growth catalyst has been the transition from regression models (ML 1.0) to deep learning recommendation models (ML 2.0), which significantly increased advertiser return on investment (ROI).
  • The company is expanding beyond mobile game user acquisition into discovery-based e-commerce and retail advertising.
• Financial profile and valuation history:
  • Went public in April 2021 at a $28 billion valuation ($600 million EBITDA), peaked at $40 billion, then collapsed 92% to a trough market cap of $3.8 billion (~$9 per share) in 2022 despite generating $1 billion in EBITDA that year.
  • Executed an aggressive capital return strategy during the drawdown, repurchasing approximately $6 billion of its own stock and retiring 20% to 25% of total shares outstanding.
  • The stock subsequently surged from $9 to $750 per share, reaching a peak market cap of $250 billion.
  • Current operations generate an industry-leading 84% to 85% EBITDA margin, with expectations to generate around $6 billion in cash.
• Operational and competitive advantages:
  • Divested internal gaming studios that were originally purchased for model training data, removing conflicts of interest with third-party game developers.
  • Successfully adapted to Apple's iOS privacy tracking restrictions through deep learning models that optimize ad relevance without tracking precise individual locations.
  • Maintained high margins due to an automated, lean operational footprint supported by key engineering hubs in Palo Alto, Beijing, and Singapore.

Takeaways

• High-Margin Cash Generation: With EBITDA margins around 84%–85%, AppLovin demonstrates immense pricing power and operational leverage in programmatic advertising.
• AI-Driven E-Commerce Expansion: The primary growth upside lies in porting its deep learning recommendation engine from mobile gaming into broader e-commerce and consumer goods advertising.
• Disciplined Capital Allocation: Management has a proven track record of counter-cyclical share repurchases, heavily buying back stock during deep market dislocations.

Meta Platforms (META)

• Meta was highlighted alongside AppLovin as a premier "discovery-based" advertising platform rather than an intent-based search platform.
  • Discovery advertising introduces consumers to products they did not previously know existed, driving discretionary transactions and expanding aggregate economic activity.
  • Platforms like Instagram have transitioned ads into high-performing content formats that drive significant retail shopping behavior.
• Insulation against AI chatbot disruption:
  • Discovery-driven ad models are substantially more insulated from conversational AI/LLM disruption than traditional search engines because consumers use social platforms for entertainment, browsing, and impulse shopping rather than specific question-answering.

Takeaways

• Durable Ad Moat: Discovery-based advertising formats remain highly resilient against large language models and chatbot interfaces, which are primarily optimized for direct lookup rather than product discovery.

Alphabet (GOOGL)

• Google's core search advertising franchise represents "bottom-of-the-funnel" intent capture, where users already know what they want to purchase and are researching where to transact.
• Competitive threats from Generative AI:
  • Bottom-of-the-funnel ad queries are directly vulnerable to conversational AI models and large language models (LLMs) that can resolve user queries and close transaction loops without traditional search result pages.
  • Emerging ad products from AI chatbot providers (such as OpenAI/ChatGPT) will compete directly for Google Search market share and intent-based ad budgets.

Takeaways

• Search Query Vulnerability: Google faces structural headwinds in its high-intent search ad business as conversational AI models increasingly become the entry point for direct consumer research and transaction execution.