Bio
Head of Growth at Anthropic, running a 40-person org structured as horizontals (growth platform, monetization) plus audience pods (Claude Code, Cowork, knowledge worker, B2B, API). Previously led growth work at Mercury and MasterClass. Public operator on the CASH framework — Claude Accelerates Sustainable Hypergrowth — Anthropic's substrate for automating the four-stage growth experimentation loop.
Operating themes
- Substrate runs experimentation; humans run alignment. The four stages of a growth experiment can be automated; cross-functional agreement cannot.
- The squeezed PM. Claude Code multiplies engineering 2–3x and pushes PM and design into the bottleneck.
- Right friction beats no friction. Keep the friction that teaches the user about the product; cut the rest.
- Quality as a growth lever. In regulated or complex flows, fixing quality outperforms conventional optimisation.
Cards
- Automate the four stages of a growth experiment; keep humans on alignment — Automate identify → build → test → analyze; keep humans on alignment [Tier A]
- Claude Code multiplies engineers 2–3x; PM and design become the bottleneck — 2–3x eng multiplier turns PM and design into the bottleneck [Tier A]
- Keep friction that helps users understand the product; cut the rest — Friction that teaches the user is a feature, not a tax [Tier A]
- At scale, PMs should up-level the why and what; not ship more features — At 20:1 eng:PM ratios, sharpen direction; don't ship the 21st feature [Tier B]
- Brand and quality are growth levers, not constraints on growth — In complex flows, fix quality before running A/B tests [Tier B]
Sources captured
- 2026-04-05 — Lenny's Podcast, "Anthropic growth, CASH, and the squeezed PM" (
raw/podcasts/amol-avasare--anthropic-growth-cash--2026-04-05.md)
External
- Lenny's Podcast: https://www.lennysnewsletter.com/podcast