Menlo's Murphy: AI Startups Must Achieve Impossible Growth, Like Anthropic

Menlo Ventures partner Matt Murphy has spent a quarter-century watching startups claw their way to profitability, surviving recessions, and occasionally achieving the kind of growth that makes spreadsheets legitimate. But none—not Netscape, not Instagram, not Stripe—prepared him for what he claims to have witnessed at Anthropic, the AI company in which Menlo led a $500 million Series funding round. According to Murphy's testimony, Anthropic somehow leaped from a $9 billion annual run rate in 2025 to $47 billion by May 2024. That's a 422% increase in annualized revenue in the span of months. Murphy, understandably gobsmacked, declared this the most explosive growth trajectory he'd encountered across three decades of internet, mobile, and cloud cycles. The bar for awe, apparently, has relocated to the stratosphere.

What makes this claim particularly theatrical is the near-total absence of context around what "revenue run rate" actually means in the Anthropic case. The company sells API access to Claude, its large language model—a product with staggering unit economics if the numbers hold, but also one whose customer base, usage patterns, and pricing remain opaque to public scrutiny. A run rate, of course, is not revenue; it is an extrapolation, a mathematical assumption that current performance will persist indefinitely. It is the financial equivalent of a junior analyst saying, "If this trend continues forever, we're rich." Anthropic's claimed run rate would place it in the revenue stratosphere occupied by pharmaceutical giants and Fortune 500 tech companies—all without the burden of actually generating audited financial statements, customer concentration disclosures, or the minor inconvenience of profitability.

Murphy's astonishment is itself worth scrutiny. A 25-year veteran of venture capital has witnessed the dot-com boom, the mobile explosion under Apple and Google, and the cloud infrastructure race that made Amazon's AWS a $80+ billion business within a decade. Yet none of those trajectories, in his recollection, resemble Anthropic's alleged ascent. One might wonder: is this a genuine moment of market disruption, or is Murphy simply observing a company whose claims have not yet collided with the gravity of verification? The venture capitalist's role is partly to marvel publicly at his own prescience, and there is no faster way to do that than to declare the rules of business fundamentally altered.

The subtext of Murphy's comments—and the broader arc of recent AI funding—is simple enough: founders of AI startups must prioritize growth at velocities that would bankrupt conventional businesses. They must build products with unit economics so favorable that a $500 million check becomes a rounding error. They must convince enterprises to retool entire workflows around their models, and do so at a pace that makes traditional SaaS onboarding look like geological time. This is not advice; it is a threat disguised as encouragement. Founders who cannot achieve Anthropic-scale growth have, implicitly, failed to understand what is different about AI.

History offers a cautionary appendix to this narrative. The last time venture capitalists unanimously agreed that growth was the only metric that mattered—that unit economics and path to profitability were secondary concerns—was approximately 2020-2021. The companies that survived that era did so not because they achieved impossible growth, but because they eventually had to grow up. Stripe, for instance, genuinely revolutionized payments, but it took seven years of quiet execution before achieving decacorn status. Anthropic, by contrast, is being asked to perform miracles within quarters, and to do so while OpenAI, Google, and Microsoft pour incomprehensible resources into AI development.

What Murphy's amazement really signals is not the arrival of a new paradigm, but the moment when venture capital's appetite for audacity outpaced its tolerance for evidence. A $47 billion revenue run rate is not a fact to be marveled at; it is a claim to be interrogated. The question is not whether Anthropic has achieved unprecedented growth, but whether its financial assertions will survive contact with reality—or whether they exist primarily in the soft, judgment-free zone of venture enthusiasm.

In the end, Murphy's 25 years of experience have taught him that markets always surprise. They simply don't usually surprise in the direction of founders achieving impossible things on unverified schedules.