AppLovin CPO on how renting AI can be a waiting game

This article translates and summarizes an episode of Silicon Valley 101’s video series featuring Ge Xiaochuan, chief product and engineering officer at AppLovin.

AppLovin chief product and engineering officer Ge Xiaochuan / Photo credit: AppLovin

AppLovin started in 2012 by helping mobile game developers find players and sell ads. The company recently built a new AI system to run its advertising tools and expand into other verticals like online shopping.

This move grew its 2025 annual revenue by 70% year on year to US$5.3 billion. The company also generated US$3.3 billion in net income for the year.

It achieved this growth without building its systems from scratch.

Ge Xiaochuan, AppLovin’s chief product and engineering officer, believes renting basic technology offers a clear advantage. By relying on outside tools, the team’s engineers can focus on helping users and expanding the business.

“Because we aren’t self-developing, we have full autonomy to choose the best models available on the market,” Ge says. He also highlights that the team stops worrying about how to build the AI and starts focusing on how to use it in its apps.

Accelerating the product pipeline

While leaving the hardest work to other companies helps make the internal team more efficient, it also creates an unusual development cycle. Engineering plans must match the predicted release schedules of outside platforms.

“We waited for them on the other end of the timeline,” Ge explains. “When Veo and Sora came out, we knew that the right moment had come. At that time, our entire product flow was already ready.”

Waiting for external tools requires discipline, but Ge argues that holding back beats chasing trends. 

While some worry that waiting gives too much power to outside companies, he dismisses this fear. Since AppLovin already relies on Google to run its daily operations, Ge sees no reason to panic.

“If we really worried about Google cutting off our access, should we also worry that we can’t use Google Cloud?” he points out.

Navigating unfamiliar markets

AppLovin’s decision to expand from games to online shopping means dealing with different customers. To survive, the team had to forget their past wins and look at the facts of the new market.

“I think the worst mistake is relying purely on past experience,” Ge notes. “At this time, it requires First Principles thinking. You have to realize it is a different product.”

Rethinking talent acquisition

Defining the cultural baseline