💡

InsightHunt

Hunt the Insights

H

Howie Liu

Co-founder & CEO

Airtable

👥 Team & Culture (1)🎯 Product Strategy (1)🔍 User Research (1)

Key Takeaways

  • 1.Software companies must be 'refounded' for the AI era; if legacy assets don't provide leverage, consider selling.
  • 2.CEOs should return to being 'IC CEOs' (Individual Contributors), acting as Chief Tastemakers by using AI tools hourly.
  • 3.Split product teams into 'Fast Thinking' (experimental, shipping weekly) and 'Slow Thinking' (scalable infrastructure) groups.
  • 4.In early AI product development, prioritize 'vibes' and open-ended exploration over rigid 'evals' (evaluations).
  • 5.The roles of PM, Designer, and Engineer are collapsing; everyone must become a hybrid builder with proficiency in all three areas.

Methodologies(3)

👥 Team & Culture

Inspired by Daniel Kahneman's concept, Airtable split its Engineering/Product/Design (EPD) organization into two distinct groups. The 'Fast Thinking' group operates like a startup to ship AI features weekly, while the 'Slow Thinking' group focuses on deliberate, long-term infrastructure and scalability.

Core Principles

  • 1.Fast Group (AI Platform): Operates with high autonomy, ships weekly, focuses on 'wow' factor and top-of-funnel excitement.
  • 2.Slow Group (Core/Infra): Focuses on deliberate planning, scalability (e.g., handling 100M+ records), and reliability.
  • 3.Complementary Growth: Fast thinking attracts new users (top of funnel); Slow thinking retains them and enables enterprise expansion.

"We have the fast thinking group... we want to just ship a bunch of new capabilities on a near weekly basis. And each of them should be truly awesome value."

#'fast#thinking'#structure
View Deep Dive →
🎯 Product Strategy

A strategic framework for founders to decide how to pivot. It asks leaders to imagine starting from scratch today with AI-native approaches and evaluate if their current assets (codebase, brand, customers) act as leverage or liability.

Core Principles

  • 1.Clean Slate Imagination: If you founded the company today with the same mission, how would you execute using a fully AI-native approach?
  • 2.Asset Valuation: Do your existing building blocks (e.g., no-code primitives) give the AI leverage, or are they legacy debt?
  • 3.The 'Sell' Criterion: If you cannot execute better with your current assets than a new startup could, you should find a buyer and start over.
  • +1 more...

"If you can't... then you should find a buyer and then if you really care about this mission, go and start the next carnation of it."

#refounding#strategy#product
View Deep Dive →
🔍 User Research

Instead of starting with rigorous evaluations (evals) or metrics, teams should start with open-ended 'play' to develop an intuition for what the model can do. Formal metrics should only be introduced after the product form factor and use cases have stabilized.

Core Principles

  • 1.Mandatory Play: Encourage teams to block out days or weeks just to 'play' with new AI models without a specific deliverable.
  • 2.Vibes Over Evals: In the 0-to-1 phase, rely on human intuition and 'vibes' to judge quality, as AI capabilities are often unknown until tested.
  • 3.Convergent Metrics: Only implement rigorous 'evals' once you have identified the cluster of useful use cases (diverge first, then converge).
  • +1 more...

"For a completely novel product experience or form factor, you should actually not start with evals and you should start with vibes."

#vibes-first#discovery#process
View Deep Dive →