Anthropic’s $1B to $19B growth run: how Claude became the fastest-growing AI product in history
TL;DR
Anthropic went from $1B to $19B ARR in 14 months by compounding product breakthroughs, not just growth hacks — Amole Evasari says the company’s 10x year-over-year pattern has held from 2023’s $0 to $100M, to 2024’s $100M to $1B, to 2025’s roughly $1B to $10B, with the Feb. 2026 $19B figure already “out of date.”
Inside Anthropic, growth is mostly managing “success disasters” — Evasari estimates 70% of his job is firefighting issues created by hypergrowth across acquisition, activation, and monetization, with only 30% on classic proactive growth work.
One of Anthropic’s sharpest growth lessons is that the right friction increases conversion — Drawing on Mercury, MasterClass, and Claude onboarding, Evasari argues that asking users who they are and what they want often beats “remove every step,” because it helps route them to the right product and aha moment.
Anthropic is already using Claude to automate parts of growth experimentation — The internal “CASH” effort (“Claude Accelerates Sustainable Hypergrowth”) can identify opportunities, build simple changes, test them, and analyze results, and Evasari says it currently performs around the level of a junior PM while already “printing money.”
AI is changing org design in a counterintuitive way: engineers got the biggest boost first, so PMs and designers are now squeezed — At Anthropic, a team that once felt like 5 engineers, 1 designer, 1 PM can now behave like 15–20 engineers, forcing more PM hiring and a rule that engineers act as “mini PMs” for projects under two engineering weeks.
Anthropic’s focus on coding and B2B wasn’t accidental—it was strategy and survival — Evasari says leadership, especially Dario Amodei, saw coding as both a huge market and a research flywheel, while Anthropic’s lack of Google/Meta distribution or OpenAI’s head start forced a narrow focus that became a superpower.
The Breakdown
How Amole got the job with a cold email
Evasari didn’t come through referral, sourcing, or even an open role—he cold-emailed Anthropic CPO Mike Krieger saying, essentially, “great product, but you badly need a growth team.” He credits years of founder-style cold outreach: tested subject lines, personal email over crowded channels like LinkedIn, brutally concise copy, and relentless follow-up until someone says stop.
The surreal reality of leading growth at a $19B ARR rocket ship
Lenny opens with the absurdity: Anthropic went from $1B ARR at the start of 2025 to $19B 14 months later, with Evasari noting even that figure is stale. Internally, he says, linear charts are “not cool”—everything is viewed on a log scale—and much of the company still feels like it’s hanging on “by the seat of our pants.”
Why growth at Anthropic is harder than it looks
Even with “the best models in the world,” Evasari calls this the hardest job he’s ever had, harder than founding a company or working in investment banking. The work looks normal on paper—acquisition, activation, monetization—but 70% of his time goes to “success disasters,” where things are growing so well that systems, flows, and teams start breaking under the strain.
Activation, memory import, and the case for good friction
Anthropic’s much-discussed ChatGPT memory import was framed as one example of solving the AI cold-start problem: helping Claude understand users faster. Evasari’s bigger point is that activation is the highest-leverage problem in AI products, and that adding friction—questions about who the user is and what they want—often improves conversion because it lets the product recommend the right experience instead of dumping everyone into the same blank box.
The Mercury onboarding lesson he brought to Claude
At Mercury, Evasari’s team spent an entire quarter ignoring metrics and just fixing the quality of a painful onboarding flow full of banking edge cases like legal, physical, and registered-agent addresses. That quality-first push ended up being one of the most impactful quarters of his career, and he says the same principle carries to Anthropic: better UX is growth, even if the flow looks “longer” from the outside.
Why Anthropic’s growth team swings bigger than most
The org is around 40 people, with horizontals like growth platform and monetization plus pods for B2B, Claude Code, API, and knowledge workers. But unlike traditional growth teams that bias toward lots of small wins, Anthropic leans much harder into bigger bets because, in Evasari’s view, AI-native products are riding exponential value creation—the product two years from now could be 100x to 1,000x more valuable than today.
Claude is starting to automate the growth loop
Anthropic’s internal CASH initiative uses Claude to identify growth opportunities, build lightweight experiments, QA against brand and quality constraints, and analyze results after launch. It’s still mostly copy edits and small UI tweaks, but Evasari says the system is already effective enough to compare to a junior PM, and the team is measuring progress week by week as more of the loop becomes automated.
PMs, engineers, and the weird new org chart AI is creating
Evasari’s read is that engineering has gotten the largest AI leverage boost so far, which means PM and design are becoming bottlenecks rather than becoming obsolete first. Anthropic’s answer is pragmatic: hire more PMs, favor highly product-minded engineers, and let engineers own projects under two engineering weeks—including talking to legal, security, and stakeholders—because in this new world, mini-PM engineers become wildly more valuable.
Focus, safety, and the culture he thinks no one can copy
Anthropic’s early obsession with coding and B2B came from leadership conviction, necessity, and a belief that coding accelerates research itself. Evasari says the company is genuinely mission-driven on safety all the way down to its public benefit corporation structure, and he ends on what he sees as the real moat: a transparent, intense culture where nobody is checked out, anyone can challenge Dario publicly in Slack, and the talent density feels like “playing for Real Madrid.”