Ben Horowitz: Success Is 17 Small Decisions

Why Hesitation Kills More Companies Than Wrong Decisions

This is Ben Horowitz - the Z in A16Z, investor in OpenAI, Cursor, Databricks - explaining leadership philosophy forged through multiple near-death company experiences. His framing cuts through a lot of conventional wisdom.

"The pilot said all plane crashes are a series of bad decisions." JFK Jr.'s crash wasn't one mistake - it was 17 decisions that were each individually defensible but compounded into catastrophe. Success works the same way. Small hard decisions lead to more small hard decisions, and eventually you get an outcome. Nobody writes about the 17 steps. They write about the neat narrative.

"The worst thing you do as a leader is hesitate." Not make the wrong decision - hesitate. The thing that causes hesitation is both options are horrible. Should we rearchitect and miss the quarter, or ship bad architecture? CEOs avoid the topic entirely. But if action is better, not deciding means the whole company gets nervous while you avoid reality.

"We went public with $2M in trailing 12-month revenue at 18 months old." Everyone knew it was insane. Wall Street Journal wrote about how stupid he was. Business Week called it "The IPO from Hell." But bankruptcy was worse. Many CEOs who hesitated on that decision went bankrupt. The psychological muscle is looking into the abyss and choosing the slightly less bad option.

The Databricks story is perfect. Six PhD students asked for $200K. Ben said no - he'd write $10M. "This company needs to build a company. You need to really go for it." The alternative was stay in school. That $200K ask would have doomed them to think too small. The best part: Ben had no idea they had a future CEO inside the company (Ali Ghodsi).

"The only value you ever add is decisions people don't like." If everyone agrees, they would have done it without you. Leadership is specifically the unpopular calls.

10 Leadership Lessons From Ben Horowitz

• 17 small decisions - Success and failure are compound effects, not single moments
• Hesitation is worst - Not wrong decisions; the delay while avoiding horrible options
• "IPO from Hell" - $2M revenue, 18 months old, but bankruptcy was worse
• Run toward fear - The psychological muscle of looking into the abyss
• $200K → $10M - Databricks needed to think like a real company, not a professor's side project
• "Managerial leverage" - If you're telling reports what to do, that's no leverage
• VP→CEO different - VPs develop people; CEOs don't know those domains well enough
• Sunk cost kills - Ability to psychologically abandon prior investment saves companies
• "Only irrational reason to start" - Not worth the money; $1.6B exit "wasn't worth the money"
• Unpopular = valuable - If everyone agrees, you added nothing

What Running Toward Pain Means in the AI Era

Leadership isn't making the right decisions - it's making any decision when both options are horrible. The companies that die aren't wrong; they're frozen. When AI compresses timelines and raises stakes, the ability to run toward pain becomes the differentiating skill.