Google's Robby Stein on scaling product decisions from startups to billions

Building a first product and maintaining one that billions depend on require fundamentally different approaches. At the upcoming TechCrunch Disrupt conference, this tension will be front-and-center in a conversation with Robby Stein, vice president of product for Google Search, who will explore how large-scale products balance moving quickly with earning trust, and pushing innovation without jeopardizing reliability.

The shift from speed to responsibility#

The instincts that propel early-stage startups—experimenting rapidly, iterating based on feedback, taking calculated risks—often collide with the demands of operating at massive scale. When changes reach billions of people, even small decisions carry global consequences. The session, titled "From MVP to Billions of Users: How Product Decisions Must Change at Scale," will examine where speed still matters and where deliberation becomes essential.

Rather than prescribing a single playbook, Stein's conversation will focus on the reasoning behind high-stakes product decisions—helping founders understand not what Google does, but why the company makes choices differently at scale.

A career spanning both worlds#

Stein's background uniquely positions him to discuss this transition. He co-founded and led Stamped, a social recommendations startup acquired by Yahoo, then spent five years at Instagram building and scaling Stories, Feed, Direct Messaging, and Reels. After leading product at AI-powered news platform Artifact, he returned to Google in 2024 to work on search and generative AI products.

This arc from founding to managing global-scale platforms means he has experienced both the freedom and constraints of each environment, offering practical perspective on when and how product approaches must evolve.

AI integration at scale#

Stein's current work touches a particularly urgent challenge: bringing new technology into established products without destabilizing them. Google Search's AI Mode surpassed 1 billion monthly active users within a year of launch, and recent updates have allowed users to move fluidly between quick AI answers and deeper follow-up conversations.

This real-world example illustrates the core tension—introducing generative AI without undermining the core search experience that users have relied on for decades. For founders integrating AI into their own products, the lesson extends beyond the technology itself to how people actually perceive and use the changes.