Sell First: Makino Masayuki on Winning Startup Strategy in the AI Era
Works Applications founder Makino Masayuki argues that selling before the product is ready — and absorbing every customer complaint — is the only path to building a great product.
- Makino’s core principle: good products don’t sell themselves — sold products get better through customer complaints and reinvestment.
- Integrated suite products are dinosaurs; specialization wins because integration multiplies codebase complexity exponentially (a 100-person-month feature becomes 1,000 in a bloated suite).
- Multi-product companies face fatal engineer attrition: top engineers can now build a rival product with 5 people, so they leave rather than rank 20th inside a large org.
- AI agents should be owned by specialized vertical vendors, not horizontal players like OpenAI or Anthropic, because only the specialist vendor can design domain-specific agent logic.
- Pay top talent above major Japanese megabank salary levels upfront — deferred comp promises are “fraud” since employees capture far less upside than founders holding equity.
- Founders must retain final product direction permanently; the moment they hand it off completely, the company starts losing to faster-moving competitors.
- Assign top 2% hires to problems nobody else has touched, with full delegation and zero expectation — 30 points out of 100 is a win on an untouched problem.
- Sell aggressively before the product is polished: promise broad functionality, collect diverse customer demands, then find the greatest-common-denominator features that reliably hit market needs.
2025-11-04 · Watch on YouTube
