Head of Growth (Anthropic): “Claude is growing itself at this point” | Amol Avasare

Episode Summary

Executive Summary: Anthropic’s head of growth, Amol Evasari, explains how the company scaled from $1B to $19B ARR in 14 months, why growth at an AI-native company is different, and how the team uses AI to automate parts of growth work. He emphasizes focus, mission, strong culture, and safety-first growth, while arguing that AI will increasingly automate experimentation, analysis, and coordination—but not human alignment.

Main Topics: Anthropic’s unprecedented growth trajectory (Priority: 5/5): The conversation opens with Anthropic’s extraordinary revenue growth, framing the company as historically fastest-growing at scale and setting up questions about how growth operates inside the org. How Amol landed the role via cold email (Priority: 4/5): Amol describes reaching Anthropic by emailing Mike Krieger after noticing the company lacked a growth team, and how that outreach reflected his long-honed cold email skills. What growth means at an AI-first company (Priority: 5/5): Amol explains that Anthropic growth still covers acquisition, activation, and monetization, but much of the work is firefighting success disasters and adapting to rapidly changing model capabilities. Activation, friction, and onboarding strategy (Priority: 5/5): A major theme is that AI products require strong activation and sometimes intentional friction so users can be matched to the right product, feature, or use case. AI automating growth work (Priority: 5/5): Amol details CACHE, an internal effort to use Claude to automate opportunity detection, feature building, testing, and analysis, with human review gradually shrinking over time. Role evolution across PM, design, and engineering (Priority: 4/5): He argues engineers are currently getting the most leverage from AI, while PMs/design are squeezed; this may increase the need for product-minded engineers and better PM leverage. Mission, safety, culture, and constraints (Priority: 5/5): Amol stresses that Anthropic’s public benefit mission, safety posture, transparency, and unusually high talent density are central to its success and long-term advantage.

Key Arguments: Anthropic’s growth is not primarily driven by growth tactics; it is driven by model quality, research strength, inference/computing, and product excellence. At AI-native companies, growth should shift toward larger bets because product value can expand exponentially and small optimizations miss the bigger opportunity. Activation is more important in AI than in traditional software because users need help discovering what the model can actually do and how to get value from it. Adding friction can improve conversion when it helps identify who the user is and routes them to the right product or workflow; friction is not always bad. Growth teams should increasingly use AI to automate parts of the experimentation loop: opportunity identification, feature creation, QA/testing, and learning analysis. Human alignment, stakeholder management, and brand/safety judgment will remain difficult to automate, especially for large cross-functional projects. PMs and designers may become more leveraged but also more strained as engineers accelerate; product-minded engineers become more valuable, and PMs should focus more on judgment and coordination than on only shipping. Anthropic’s focus on safety and public benefit is not just branding; it is embedded in its structure and culture, and the company is willing to sacrifice short-term metrics for long-term trust. Constraints can create clarity and force better decisions, both in business strategy and in personal resilience. A company’s culture and internal transparency mechanisms can be data for future AI agents, improving organizational coordination and decision-making over time.

Data Points: ARR growth: $1B to $19B in 14 months - Anthropic’s revenue trajectory as discussed at the start of the episode Mid-2025 ARR: ~$4B - Referenced as an intermediate milestone in Anthropic’s growth End-2025 ARR: ~$9B then ~$19B - Speaker cites rapid revenue doubling and later numbers that exceed prior public estimates Revenue growth rate: 10x year over year - Describes Anthropic’s recurring growth pattern 2023 revenue: 0 to $100M - Historical progression shared by Amol 2024 revenue: $100M to $1B - Historical progression shared by Amol Growth team size: ~40 people - Approximate size of Anthropic’s growth organization Growth time allocation: ~70% success disasters / 30% proactive growth - Amol’s breakdown of how his time is spent Product value outlook: 100x to 1000x in two years - How Amol describes future value from AI products CACHE automation maturity: Early; started a couple months ago - Anthropic’s internal Claude-powered growth automation effort Human-reviewed experiment performance: Comparable to a junior PM - Current quality level of automated growth outputs in CACHE Project threshold: 2 engineering weeks - If a project is shorter than this, engineers act as the de facto PMs Mercury onboarding impact: Single most impactful quarter of Amol’s growth career - Result of focusing on onboarding quality improvements Brain injury recovery: 9 months off work - Amol’s traumatic brain injury recovery timeline Re-injury recovery: 2 months off work - He was re-injured shortly after returning to work at Mercury

Pivotal Quotes: "It's the hardest job I've had in my life to come into Anthropic." — Amol Evasari: Describing the complexity of leading growth at an AI company that is scaling extremely fast "I think that the product value that we will deliver in two years' time is probably like 1,000x what it is today." — Amol Evasari: Explaining why Anthropic’s growth team should favor larger bets over micro-optimizations "If the primary value that your product delivers is underpinned by AI as a central element of it, then I think you should operate this way." — Amol Evasari: Clarifying when AI-native companies should shift toward larger growth experiments

Implications: AI-native companies will increasingly automate growth work and redesign PM/eng/design boundaries. The winners will combine technical leverage, strong activation, and mission-driven trust while using AI to scale judgment—not just execution.