The Twenty Minute VCAnthropic Raises $13BN & OpenAI Buys Statsig for $1.1BN All Stock
CHAPTERS
Anthropic’s $13B raise at ~$183B post: does the valuation math work?
The group breaks down Anthropic’s massive round, focusing less on the headline number and more on what the implied forward revenue multiples look like if current growth persists. Rory walks through ARR-to-GAAP revenue logic and shows how the deal can look ‘reasonable’ under aggressive but not impossible growth assumptions.
Why top growth funds feel forced to own LLM winners (and why rounds get oversubscribed)
They discuss the institutional pressure on large growth investors to own exposure to the category-defining model companies. Oversubscription isn’t just hype—it’s driven by career risk and portfolio construction logic when LLM leaders become the dominant compounding assets.
Canva’s secondary round mechanics: choosing investors, loyalty, and IPO cornerstones
Cliff explains how Canva thinks about investor selection when demand overwhelms supply, especially in a mostly-secondary context. Instead of maximizing price, the goal is building long-duration relationships with investors who could anchor an eventual IPO.
Managing employee expectations after valuation whiplash (50x → 20x → 10x)
The conversation turns to internal communication when market comps swing wildly. Cliff shares how Canva contextualizes public comps and emphasizes controllables—growth, margins, and customer value—over fluctuating market marks.
Early-adopter AI revenue vs mainstream distribution: crossing the chasm and renewals risk
Cliff warns that many AI products may be benefiting from early-adopter pull-forward and hype, which can mask weak year-two retention. The group debates whether AI companies can transition from $50–100M to $1B+ sustainably, where distribution becomes decisive.
Canva’s AI approach: ‘workhorses not gimmicks’ + internal tool consolidation
Cliff describes Canva’s AI philosophy—integrating durable AI capabilities into core workflows rather than shipping novelty features. Internally, Canva lets teams experiment broadly (multiple coding tools/LLMs) and expects consolidation once winners emerge.
OpenAI buying Statsig for $1.1B (stock): cheap, or perfectly engineered?
They analyze the all-stock acquisition of Statsig (~$75M ARR) at the same valuation as its prior round. The panel frames it as a deal that satisfies most constituencies—founders get an upgrade, late-stage investors preserve price, and buyers add a strong team—though some angels may dislike the lack of markup.
Zuck’s $14B Scale bet: predictable mess, talent churn, and looming write-down risk
The group argues the Scale/Meta structure was inherently messy—both on human dynamics and on whether the remaining Scale asset can justify a $14B mark. Rory predicts auditors may eventually force a write-down as cash exits and key talent migrates, leaving an ‘empty husk.’
Lovable at $4B and Vercel at $9B: step-up rounds, FOMO, and ‘validation pricing’
They debate whether rapid repricings reflect genuine new information or market psychology. Harry outlines three drivers—new info, mispricing, and validation—while Cliff argues that if a company can be $20–$50B, the difference between $2B and $4B may not matter much for long-term outcomes.
Is B2B SaaS ‘back’? Public comps rally as AI tailwinds emerge
The panel reviews strong earnings reactions across Snowflake, MongoDB, Okta, Zoom, and others. They argue it’s less a full renaissance and more a reset of expectations—when sentiment is too negative, modest re-acceleration can drive dramatic stock moves, especially for category leaders with distribution.
Canva’s IPO thinking: why go public when you’re profitable with tons of cash?
Cliff lays out the case for IPO primarily as an employee (and broader stakeholder) liquidity solution, plus access to deeper pools of capital and credibility. They also discuss direct listings, concluding that while logically appealing, founders may prefer the safer, proven path on the ‘biggest day’ of their company’s life.
Jensen’s $3–$4T AI infrastructure claim: ROI math vs ‘it’s still early’ optimism
They pressure-test whether trillions in AI capex can be justified by downstream profits. Rory is skeptical on macro ROI math, while Jason and Cliff argue AI penetration is still tiny and usage could scale massively—though compute optimization, on-device inference, and model distillation will shape outcomes.
AI compute as a margin tax: Notion’s 10% gross margin hit, Canva’s costs, and new pricing models
They explore whether AI will structurally compress SaaS margins via GPU/inference costs. Cliff confirms Canva can hit ~10%+ of revenue in AI-related costs today (especially when training models), but expects efficiency gains; he also details Canva’s move toward unified AI credits and hybrid seat + consumption pricing.
Distribution shifts: SEO vs ‘LLM SEO,’ Canva’s ChatGPT referrals, and what breaks media sites
They discuss the decline of traditional SEO for some businesses and the rise of LLM-driven discovery. Cliff claims Canva is a top productivity app within ChatGPT and a top-referred domain, driven by deliberate optimization work; Rory notes this is beneficial for products, but existential for content-only publishers.
The VC regret pathway: buying later, doubling down on winners, and the value of time-based data
They close with a candid discussion on missed investments and the psychology of paying higher prices later. Rory argues that two positive data points over time can be far more informative than a single early snapshot—so ‘pay the tax’ and invest if the thesis is now proven; Cliff adds that follow-on discipline often separates top outcomes.
