Key Takeaways
- Ben Horowitz advised Ali Ghodsi against selling Databricks, emphasizing its high potential.
- Ali Ghodsi became CEO of Databricks in 2016 during a crisis, leading it to a $100 billion valuation.
- Effective leadership involves 'flying low,' understanding operational details, and inspiring teams to win.
- Acquisition success relies on team, culture, product fit, and customer excitement, not just financials.
- Setting aggressive goals and competitive P95th percentile compensation drives talent attraction and motivation.
- Optimal startup timing is critical; Databricks' 2013 founding year was optimal for market conditions.
Deep Dive
- The board was initially skeptical of Ali Ghodsi's appointment as CEO of Databricks in 2016.
- Ben Horowitz secured a one-year deal for Ghodsi to prove himself as CEO despite initial board skepticism.
- Ali Ghodsi, from his internal perspective, knew Databricks' issues and acknowledged Ben Horowitz's mentorship.
- Databricks faced a significant business challenge as cloud vendors offered Apache Spark, its open-source project, freely.
- Effective feedback delivery and 'Radical Candor' were discussed, emphasizing framing criticism as helpful.
- Frequent, consistent feedback desensitizes individuals and encourages receptiveness, unlike annual reviews.
- Leadership's own work ethic, including 24/7 dedication, is crucial for scaling a high-intensity culture.
- To vet for hardworking employees, references should be asked about willingness to work late or under pressure.
- A leader's role is crucial in making teams feel they are winning, enabling higher expectations.
- The unique feeling of turning a losing situation into a winning one is difficult to replicate post-success.
- CEOs must 'fly low,' engaging directly with employees to understand operational details.
- Direct CEO engagement is crucial for accurate information, as knowledge resides with individual contributors and customers.
- A CEO cannot address all departments equally, necessitating prioritized deep dives into critical areas.
- Meeting frequency and attention should vary based on individual or department importance, avoiding uniform structures.
- The discussion covered Databricks' 2017 deal with Microsoft, which began after Ben Horowitz spoke with Satya Nadella.
- Securing significant financial commitment from Microsoft was strategic to ensure their sustained engagement.
- Large companies like Microsoft need substantial pre-commitments in deals to ensure follow-through.
- Databricks convinced Microsoft by emphasizing their hunger and Microsoft's product gap against AWS.
- Securing large deals like the Microsoft partnership required persistence and internal influence during cultural shifts.
- Databricks prioritizes acquisitions based on team and cultural fit, aiming for collaborative 'co-founder' relationships.
- Product development emphasizes a customer-centric approach, prioritizing excitement and integration needs over financials.
- Multiple product architectures introduce complexity, leading to inefficiencies in sales and support, and customer alienation.
- Financial engineering and short-term revenue boosts from acquisitions can degrade long-term product and brand.
- Acquisitions risk diluting a company's talent and culture if the acquired employee base is not high-quality.
- Ben Horowitz encouraged Databricks to aim higher, comparing its potential to Oracle in the cloud.
- Horowitz pushed Ali Ghodsi to consider Databricks becoming as large as 'FANG' companies in 2017.
- Databricks recalculated compensation, deciding to pay at the P95th percentile for engineers, informing their strategy.
- Aggressive goal-setting, though seemingly outlandish, was a significant motivator for Databricks.
- Ben Horowitz advised Ali Ghodsi against selling Databricks for $6 billion, emphasizing the rare market opportunity.
- Ali Ghodsi decided against selling, influenced by Horowitz's perspective on seizing transformative ventures.
- The intense pressure on young professionals in the current AI boom creates preoccupation with starting companies and valuations.
- CEO mentorship and opportunities for career development are significant for attracting early-career talent to smaller companies.
- Databricks' founding year, 2013, was crucial due to mature market conditions and technology availability.
- Starting Databricks in 2012 or 2014 would likely have led to failure due to immature market or being overtaken by hyperscalers.
- Databricks considered a pivot in 2015 from product-led growth to B2B enterprise sales due to initial product-led growth strategy failure.
- Hiring Ron, a sales expert without a PhD, was an uncomfortable but crucial decision for the technically-minded executive team.