Why We Fight for Berkeley: The True Cost of Lost Startups
In January 2015, Databricks made a decision that seemed small at the time. The company had grown to 46 people in just two years, and the founders decided to move their headquarters from Berkeley to San Francisco. It was a rational move. San Francisco offered more office space, easier access to customers and investors, and a proven tech ecosystem. For Databricks, it made sense.
Eleven years later, that decision looks very different.
Today, Databricks is worth $134 billion and employs 15,335 people globally. It’s one of the most successful startups ever built by Berkeley researchers. And it’s in San Francisco, not in Berkeley.
I’ve spent the last decade building and advising deep tech companies, and I’ve watched this pattern repeat over and over. A Berkeley team launches a company. They find early traction. They raise funding. And then they leave. Sometimes it’s for a bigger city. Sometimes it’s for proximity to customers. Sometimes it’s because they couldn’t find affordable office space or navigate the regulatory environment. The reasons vary, but the result is the same: Berkeley loses the company it incubated.
The economic consequences are staggering.
Using data from the 2026 UC Berkeley Haas Institute report on the startup ecosystem, I ran the numbers. If Databricks had kept just 10% of its current workforce in Berkeley, the city would have $331 million in annual payroll, $603 million in direct economic value, and over 1,500 high-wage jobs. When you factor in the regional ripple effects (the vendors that would supply the company, the restaurants where employees would eat, the housing they’d occupy), you’re talking about 3,145 total jobs and nearly $1 billion in economic impact across Alameda County.
One company. Ten percent of its workforce. Nearly $1 billion in regional economic value.
That’s why we built the Berkeley Gateway Accelerator.
BGA is rooted in a simple conviction: Berkeley’s greatest competitive advantage is our unique ecosystem. We have world-class research institutions (UC Berkeley, LBNL), a thriving arts and culture community, and access to extraordinary talent. Our job is to make sure the deep tech companies built from that talent and research don’t have to leave to scale.
That’s why we started BGA. The economics of startup relocation are brutal for cities but straightforward for founders. When you’re small and uncertain, moving is cheap. You’re just moving people. But the long-term cost to the city is enormous. Every year a company grows elsewhere instead of here is a year of foregone tax revenue, foregone employment, foregone spillover effects into the broader economy.
Our model is direct: create a residency program where ambitious deep tech teams can build at scale from Berkeley. Provide the physical space in downtown Berkeley, the investor connections, the mentorship, and the community they need to succeed without leaving. We’re not waiting for the city to solve every problem. We’re building the infrastructure ourselves: the space, the ecosystem, the economic foundation that makes it rational for companies to stay.
BGA exists because we believe Berkeley’s future depends on it. We believe that the next Databricks, the next transformative deep tech company, shouldn’t have to choose between staying home and succeeding. We believe you can do both.
The Databricks analysis provides strong evidence, the economics are clear. The only question is whether we have the will to build what's needed before the next breakout company leaves.
For a detailed breakdown of all calculations and methodology, download the full analysis below.
