Summary
Aliisa Rosenthal, now a partner at Acrew Capital, spent four years at OpenAI helping grow its enterprise revenue from a few million dollars to several billion — and this conference talk distills the hard lessons from that journey into a practical go-to-market playbook for AI startup founders.
The central argument is to invert the traditional sales-first approach: build automation before headcount. Rosenthal describes how OpenAI launched ChatGPT Enterprise at $60 per user per month before building a self-serve option, then watched a January 2024 self-serve launch rapidly cannibalize the enterprise business — because most buyers simply didn’t want a heavy sales process. The lesson: launch self-serve first, observe where customers get stuck, and only then build the premium tier. She also covers how pricing set when OpenAI was the only AI enterprise product on the market had to be revised once Copilot, Gemini, and Anthropic entered at lower price points, and how shifting to a license-plus-usage model dramatically lowered the barrier to entry and drove adoption.
The talk includes practical tactics for early-stage companies: using tools like Clay to manage overwhelming inbound volume (OpenAI hit 10,000 inbound inquiries per day at peak with only five sales reps), automating security questionnaires and NDA signing through trust portals to prevent deals from stalling, and structuring contracts with opt-out clauses to reduce proof-of-concept friction. Rosenthal’s perspective — shaped by watching the AI enterprise market form in real time — makes this a useful reference for founders deciding how to sequence product, pricing, and sales hiring.
📺 Source: AI Engineer · Published August 26, 2026
🏷️ Format: Keynote Launch
