No Priors Live: Building Durable Software in the AI Age with MongoDB President & CEO CJ Desai
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
In this live recording from MongoDB.local SF, CJ Desai, CEO and President of MongoDB, discusses the future of software in an AI-driven world with host Sarah Guo. (00:58) The conversation explores why platforms outlast products, the challenges facing incumbent software vendors, and what constitutes a true "moat" in the age of generative AI. (04:56) CJ argues that while products can be replaced, platforms become sticky through deep customer integrations and multi-product usage. (21:55) He also addresses the current "bear thesis" on SaaS, explaining why Fortune 500 companies are still moving slowly on AI adoption despite the technology's potential.
- Main Theme: The distinction between replaceable products and durable platforms, and how enterprise software companies can navigate the AI transition while maintaining competitive advantages through customer relationships and integrated ecosystems.
Speakers
CJ Desai
CJ Desai is the CEO and President of MongoDB, having recently taken the helm of the $10+ billion database platform company. He previously held senior leadership roles at Cloudflare and ServiceNow, where he gained extensive experience in building and scaling enterprise software platforms. CJ began his career at Oracle Corporation, where he learned foundational principles about database platforms and enterprise software scaling.
Sarah Guo
Sarah Guo is a venture investor and co-host of the No Priors podcast, focusing on the future of software and AI technologies. She has extensive experience in the technology investment ecosystem and regularly analyzes market trends in enterprise software and artificial intelligence.
Key Takeaways
Platforms Are Sticky, Products Are Replaceable
CJ emphasizes that the fundamental difference between successful and struggling software companies lies in whether they position themselves as platforms or mere products. (04:56) He explains that while individual software tools can be easily swapped out, platforms become deeply embedded in customer infrastructure through multiple product integrations and system dependencies. A compelling example he shares involves a London bank that built 300 critical applications on MongoDB out of their total 9,000 applications - demonstrating how platform stickiness increases with usage. (08:59) This principle explains why only single-digit companies have achieved over $10 billion in pure-play software revenue - because platforms are rare and difficult to replicate.
Speed and Learning During Technology Transitions
During major technology shifts like the internet age, mobile revolution, or current AI transformation, the companies that survive and thrive are those that pivot quickly and learn fast. (02:21) CJ argues that having a "moat" isn't enough if you fall behind during technological transitions. Companies must continuously innovate and show reacceleration of growth using new technologies like AI. (13:39) The key is not just to innovate more, but to innovate in ways that directly translate to selling more - if you're innovating but not growing sales, investors and customers will question your future relevance.
Customer Relationships Require Deep Understanding and Presence
True customer relationships in enterprise software go beyond just having a sales channel - they require deep understanding of customer problems and consistent engagement. (29:09) CJ learned from former Symantec CEO John Thompson that product leaders cannot be effective unless they speak to customers constantly, not just to ask how to serve better, but to understand their broader pain points and see around corners. (30:01) This customer intimacy enables better product decisions and helps identify platform expansion opportunities, like discovering that a retailer using MongoDB for e-commerce wasn't aware of additional capabilities like vector search.
Enterprise AI Adoption Is Still In Early Stages
Despite the hype around AI, Fortune 500 companies are moving slowly and getting mixed results from their AI investments. (22:33) CJ reports that large enterprises have seen unclear value from office productivity AI tools and are still experimenting with customer support applications, questioning whether AI-native solutions should complement or replace their existing systems of record. However, coding assistance showed breakthrough adoption in 2024, with very positive feedback on improving innovation velocity and security. (23:38) This suggests that AI adoption will vary significantly by use case and application type.
The "And" vs "Or" Decision Framework for New Technologies
When evaluating new AI-native companies, enterprise leaders are asking a critical question: should this be an "and" (addition to existing systems) or an "or" (replacement for current solutions). (24:18) CJ advocates for the "or" approach, stating he'll pay attention to vendors who can completely replace existing systems while being cheaper, faster, and better with disruptive pricing models. (25:14) This mindset represents a willingness to ignore sunk costs and embrace transformation rather than just incremental productivity gains, positioning organizations to be "AI-first" rather than just AI-enhanced.
Statistics & Facts
- Only single-digit companies worldwide have achieved over $10 billion in pure-play software revenue, despite the software industry being around for decades with many brilliant people building companies. (07:36) CJ uses this statistic to illustrate how rare true platforms are compared to replaceable products.
- A London bank CTO revealed they have built 300 critical applications on MongoDB out of their total 9,000 applications in their infrastructure. (08:59) This demonstrates the scale of enterprise application environments and how platforms can become deeply embedded while still having massive expansion opportunities.
- Oracle Corporation will celebrate its 50th anniversary in about a year and a half, while MongoDB was created in 2007, making it only 18 years old. (16:34) CJ uses this timeline to show there's still a large TAM to be disrupted in the database market, with MongoDB being the primary disruptive force in a decades-old industry.
