Dec 18, 2014 · 35m · mad
Mike Olson, Cloudera // The Cloudera Story (Hosted by FirstMark Capital)
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this Data Driven NYC talk hosted by FirstMark Capital's Matt Turck, Cloudera co-founder and Chairman Mike Olson discusses the founding and evolution of Cloudera, open-source monetization strategies, and the technological landscape of big data. He shares insights on massive strategic investments, navigating enterprise adoption, and key opportunities for modern tech entrepreneurs.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Matt holds 13.9% of the talking time here. How this is scored →
Matt as informed peer 2.6 Guest teaching 3.9 Guest disagreement 1.3 Matt pushing back 1.0
speaking balance: gold is Matt, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 21:22 Rejecting the Oracle comparison
The guest forcefully rejects the host's suggestion that Cloudera aims to be the 'new Oracle', explaining that telling IT buyers you want to be Oracle goes over terribly.
Hardest push from Matt ▶ 8:14 Enterprise adoption reality checkThe host presses past market hype to directly question whether any enterprise actually runs Hadoop core across their entire business operations.
Biggest teaching moment ▶ 19:58 Corporate governance and takeover defenseThe guest educates the host on corporate M&A defense mechanics, detailing how Intel's 18% equity stake blocks potential hostile acquirers like IBM who would need 100%.
Matt holds his own ▶ 16:55 Challenging the $900M mega-round vs IPOThe host demonstrates sharp financial awareness by citing exact fundraising figures ($900M total, $740M Intel) and pressing why Cloudera took strategic money instead of pursuing an IPO.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Matt as informed peer | Guest teaching | Guest disagreement | Matt pushing back | Why |
|---|---|---|---|---|---|---|
| Welcome and Informal Event Opening | 1 | 2 | 1 | 0 | The host opens with light banter about meeting the guest in Ireland and asks a standard opening question about his career path from Berkeley. The guest provides a friendly background story spanning BerkeleyDB, Sleepycat, and Oracle. | |
| The Founding Insights Behind Cloudera | 1 | 3 | 1 | 0 | The host asks what insight led to the founding of Cloudera. The guest explains his transition away from traditional relational databases and how meeting early Hadoop pioneers revealed a whole new scale-out paradigm. | |
| Navigating the Big Data Hype Cycle | 4 | 4 | 2 | 3 | The host probes where Hadoop sits in the hype cycle and directly challenges whether any company actually runs Hadoop across the entire enterprise. The guest jokes about market hype before clarifying that Hadoop complements rather than replaces transactional OLTP and OLAP databases. | |
| Architectural Evolution of the Hadoop Ecosystem | 3 | 5 | 2 | 1 | The host asks about technological evolution (Spark, real-time) and Cloudera's positioning relative to competitors like MapR and Hortonworks. The guest demurs on participating in vendor slap fights and educates the audience on open source curation and platform ecosystem strategy. | |
| The $900M Funding Round and Intel Investment | 4 | 6 | 1 | 2 | The host cites the massive $900M funding round and asks why Cloudera chose an Intel strategic investment over an IPO like Hortonworks. The guest explains corporate governance mechanics and how Intel's 18% stake acts as a defense against hostile takeovers by legacy giants like IBM. | |
| Open Source vs. Hybrid Business Models | 4 | 5 | 3 | 2 | The host asks if Cloudera aims to be the new Oracle and presses on their hybrid open/closed source business model. The guest rejects the new Oracle framing, explaining that pure open source leads to severe price compression while hybrid models sustain ongoing platform innovation. | |
| Reflections on Startup Velocity and Competition | 2 | 4 | 0 | 0 | The host asks about startup velocity and where new data startups should build today. The guest offers detailed strategic advice, explaining why building application-layer software on top of the platform offers higher margins than starting another platform company. | |
| Audience Q&A: Enterprise IP and M&A Strategy | 1 | 3 | 1 | 0 | Industry analyst Tony Baer asks how Cloudera maintains differentiation through proprietary IP beyond cluster management. The guest explains their M&A strategy around security and proprietary tooling where open source projects do not compete. | |
| Audience Q&A: Reaction to the Hortonworks IPO | 3 | 3 | 1 | 1 | The host asks how the guest felt emotionally about competitor Hortonworks going public first. The guest offers a gracious response, framing the successful IPO as a strong public market validation for the entire big data sector. |
Statements from this episode (14)
Sleepycat Software operated for eight years without venture capital before Oracle acquisition
“Sleepy Cat was the company that commercialized Berkeley DB in 98, and we ran that company entirely independent, no venture capital the old-fashioned way. We would make money and then we would spend it for eight years. We sold the company to Oracle in 2006”
Hadoop was designed for new data problems, not relational database issues
“What we didn't understand at the time, and it's been a pretty common feeling, is Hadoop wasn't built to solve the problem we'd been solving with relational databases. It was designed to solve a new problem, and it turned out that new problem was going to be ve…”
Hadoop's shared-nothing architecture does not easily translate to OLAP or OLTP workloads
“Hadoop, this big scale-out, shared-nothing architecture, is good at much, but that architecture doesn't easily translate into OLAP or OLTP workloads.”
Mike Olson predicts MapReduce compute cycles in Hadoop clusters will approach zero
“I think the percentage of cycles spent on MapReduce in Hadoop clusters generally is going to asymptotically approach zero. That's not because there will be less MapReduce happening, but because there will be so much of the other stuff happening.”
Cloudera uses proprietary management tools to prevent commoditization by large vendors
“Because we've got some management monitoring administrative tools That the big, well-capitalized vendors can't simply pick up and use against us. We've got a reason for customers to come and talk to us. It's not really about lock-in. It's about lock-out for IB…”
Intel acquired an 18 percent equity stake in Cloudera for $740 million
“Intel spent seven hundred forty million dollars and acquired an 18% stake in Cloudera.”
Big data market has the potential to surpass relational database market size
“I think that there's an opportunity for the big data market to be much bigger than the relational database market was, right?”
Only hybrid open-source models generate sufficient margins to fund platform innovation
“I am convinced that Only a hybrid company can basically generate the revenues and the margins that allow us to invest forward in innovation in the platform.”
IBM began offering open-source alternatives within 2.5 years of Cloudera's founding
“At two and a half years, IBM was looking at what we were doing and offering their own open source alternatives, so.”
Majority of big data market spend will go to applications, not platforms
“The lion's share of the money spent in the market Is going to be for apps, not for the platform.”
Annual spend on relational database tools is double the underlying platform spend
“Seventy billion annually is spent on services and applications and tools on top, right? So more than twice as big as the platform play is The tools play.”
Cloudera refuses to acquire companies or products that compete with open source
“So in general, we won't acquire stuff that competes with existing open source, because that would be crazy, right?”