Search for a command to run...

Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
In this episode of 20VC, Harry Stebbings, Rory Driscoll, and Jason Lemkin dive deep into the week's biggest tech news. The discussion centers on Brex's $5.15 billion acquisition by Capital One, exploring whether this represents a great outcome for founders versus the disappointment of selling below their 2021 peak valuation of $12 billion. (04:45) The conversation expands to cover the broader implications for competitors like Ramp, TikTok's complex ownership deal, Anthropic's higher-than-expected inference costs, and OpenEvidence's massive $12 billion funding round.
Host and founder of 20VC, one of the world's largest venture capital podcasts. Harry has built 20VC into a major platform interviewing top investors and founders, and has become a prominent voice in the venture capital community.
Experienced venture capitalist and tech industry analyst known for his sharp insights on market dynamics and company valuations. He brings deep expertise in evaluating both private and public market opportunities.
Founder and managing director of SaaStr, a leading SaaS community and fund. Jason previously founded EchoSign (sold to Adobe) and is recognized as one of the foremost experts on SaaS business models and growth strategies.
Jason introduces the concept of "hubristic financing" - the phenomenon where companies raise at extremely high valuations during peak market conditions, then face disappointment when they exit at lower multiples. (06:09) The Brex case exemplifies this: raising $1 billion at a $12 billion valuation in 2021, then selling for $5.15 billion in 2025. While the outcome is objectively great (building something from nothing to $5 billion), it feels disappointing relative to peak expectations. The key insight is that founders had no choice but to raise at market prices when they needed capital, making this an unavoidable cost of doing business in competitive markets.
The discussion reveals a critical challenge facing many B2B companies: inference costs are rising faster than expected, creating a potential "final nail in the coffin" for companies that aren't hyper-growth. (23:54) Jason describes a scenario where mature B2B companies with 50-200 million ARR finally build competitive AI agents, only to discover they can't afford the $20-50 million annual inference costs needed to compete with well-funded AI-first companies. This creates an impossible situation where companies are told to be profitable, then build AI capabilities, but can't fund the compute needed to be competitive.
The contrast between EquipmentShare's successful IPO (33% pop at $8 billion market cap) and Wealthfront's struggling public performance highlights that $3 billion appears to be the threshold for successful public offerings. (54:53) Companies below this scale face liquidity challenges and difficulty attracting top talent. The market clearly distinguishes between companies that have achieved true scale and those that remain "subscale," with the latter facing ongoing challenges in public markets regardless of their fundamental business quality.
Andreessen Horowitz's report claiming two-thirds of private AI revenue comes from their portfolio companies signals a fundamental shift in venture capital toward institutional scale. (45:47) The discussion explores whether venture is finally becoming a true asset class rather than a "weird niche of PE." However, this raises questions about what happens during generational transitions at iconic firms, as the success may be more tied to individual partners than institutional processes. The parallel to investment banking's evolution from partnerships to institutions provides a framework for understanding this transformation.
While AI gets most attention as a threat to SaaS, Jason identifies multiple other structural challenges: seat contractions (companies hiring fewer people), aggressive price increases crowding out upselling opportunities, and budget shifts toward AI initiatives. (70:54) Companies like Shopify have kept headcount flat for three years while growing 40%, and SaaS prices have increased 40% over recent years. These combined pressures mean that even building great AI agents may not be enough for traditional SaaS companies to regain their historical growth rates of 120-140% net revenue retention.