20VC: Leading Anthropic's First Ever Round | Will Open Source Threaten Anthropic's Business | Do Margins Matter in a World of AI | Why Triple, Triple, Double, Double is Not Good Enough Today | Why Series A is Hard Today with Matt Murphy @ Menlo

What It Covers

Matt Murphy of Menlo Ventures details how he led Anthropic's first round at a $4B valuation with a $10M check, later orchestrating a $500M+ SPV — the firm's first ever. He covers open source threats to frontier models, why Series A is the hardest insertion point today, and how venture return math has fundamentally changed.

Key Questions Answered

• •Ownership vs. Entry: Menlo passed on 1% positions in Eleven Labs, StarCloud, and Deal — all of which would have returned significant capital. Murphy's framework shift: being in a winner at 1% ownership outperforms owning 20% of a company exiting at $300–500M. In today's outlier-driven market, getting into the right company at any ownership level beats missing it entirely over ownership thresholds.
• •SPV Strategy for Concentration: When a fund's per-company cap (e.g., $100M in a $1B fund) is insufficient for a breakout company, SPVs allow continued offense without breaching fund mandates. Menlo's first-ever SPV exceeded $500M for Anthropic's follow-on round. The trigger: LP enthusiasm after an Anthropic executive presentation convinced the partnership to sign a term sheet within two weeks.
• •Series A Barbell Approach: Series A is the worst risk-adjusted insertion point today — companies show $1–3M ARR at $200–400M valuations with unproven PMF. Menlo's response: a barbell strategy of seed checks up to $8M (tripled from $3M) deployed quickly by three partners, combined with late-stage entry into clear category leaders above $10M ARR where winner dynamics are established.
• •Open Source Won't Displace Frontier Models: Application companies using Anthropic report measurably higher customer retention and user engagement versus open source alternatives, making the cost savings insufficient justification to switch. Murphy's framework: companies will run hybrid stacks — roughly 50% frontier model, 50% open source — with intelligent routing layers like OpenRouter optimizing across price, reasoning, latency, and performance at scale.
• •Gross Margin Path Matters More Than Current Margins: Many hypergrowth AI application companies currently run 20–30% gross margins due to inference costs, but the investable question is whether a credible path to 60–70% exists. Murphy's diligence framework: can the company reduce model dependency through open source integration, proprietary fine-tuned models on their own data, or multi-model routing to structurally improve unit economics over 18–24 months?
• •Relationship Timing Determines Deal Outcomes: The most common reason Menlo loses competitive deals is arriving weeks before a round closes while a competitor has a year-long relationship with the founder. Murphy's counter-strategy: Menlo's Anthology Fund deploys $100K–$1M checks into 50+ seed-stage companies, creating cap table presence that makes the firm 10x more likely to lead or participate meaningfully in subsequent rounds when breakout signals emerge.