Ben Horowitz Offered Databricks $10M Instead of $200K
by Ben Horowitz on September 11, 2025
Databricks' Early Funding: The Power of Thinking Beyond Incremental Growth
Situation
- In the early days of Databricks, six PhD students and Professor Jan Stoiker approached Ben Horowitz seeking a modest $200,000 investment
- The team had developed Spark, a technology competing with Hadoop, which already had well-funded companies pursuing it
- Spark was open source, creating time pressure as the technology was publicly available
- The founding team came from an academic background, where a $50 million outcome would be considered a significant success
Actions
- Ben Horowitz refused to write a $200,000 check, which would have enabled incremental progress
- Instead, he offered $10 million - 50x what they requested - to force a mindset shift
- He challenged them to "build a company" and "really go for it" rather than pursue a smaller vision
- Horowitz recognized they needed substantial capital to compete against well-funded Hadoop companies
- He pushed them to make a binary decision: either commit fully to building a significant company or stay in academia
Results
- Databricks accepted the larger investment and built a substantial company
- Ali Ghodsi, initially VP of Engineering, eventually became CEO (though this wasn't planned at investment time)
- The company successfully competed against Hadoop and became a major player in the data infrastructure space
- Databricks has grown to become one of the most valuable private technology companies
- Horowitz considers this "very good luck" that they had someone like Ali who could become CEO
Key Lessons
- Challenge founders to think bigger: Sometimes entrepreneurs, especially from academic backgrounds, set their sights too low based on what seems reasonable rather than what's possible
- Match funding to competitive reality: The investment size should reflect the competitive landscape, not just immediate needs
- Force binary decisions: By offering much more capital than requested, Horowitz created a forcing function that required the founders to commit fully to building a substantial company
- Recognize when incremental funding will fail: In competitive technology spaces with open-source components, moving quickly with substantial resources can be essential
- Look beyond immediate needs: Great investors see potential that founders themselves might not yet recognize
- Binary commitment matters: The "stay in school or build a real company" framing eliminated the middle ground of half-measures
- Academic success metrics differ from startup metrics: What constitutes a "win" in academia ($50M outcome) may be too small a vision for a venture-backed technology company
The Databricks story demonstrates how the right investor can reshape a founding team's ambition level and help them build something far beyond their initial vision by forcing them to make a clear choice between incremental progress and transformational commitment.