Airtable's Shift from Feature Teams to Business Units

The Fast Thinking vs. Slow Thinking Organizational Model for AI Transformation

Howie Liu, CEO of Airtable, restructured his company to accelerate AI adoption by creating two distinct organizational groups with different operating modes and expectations. This model helps established companies balance rapid innovation with infrastructure stability.

The Fast Thinking Group (AI Platform)

• Ships new capabilities on a near-weekly basis
• Focuses on creating "jaw-dropping" value with each release
• Operates with high autonomy and minimal process
• Prioritizes experimentation and rapid iteration over perfect planning
• Creates top-of-funnel excitement and inspires new use cases
• Attracts users through experiential value rather than sales pitches

The Slow Thinking Group

• Works on deliberate, premeditated bets requiring longer timelines
• Handles complex infrastructure that can't be shipped in hacky prototypes
• Builds systems that can handle massive scale (e.g., databases with 100M+ records)
• Enables initial adoption seeds to "sprout and grow into larger deployments"
• Provides the foundation that makes fast-moving innovations durable

Why This Model Works

• Complements rather than competes: Fast thinking creates excitement and adoption; slow thinking enables retention and expansion
• Addresses the challenge of AI-native companies that get "tourist traffic" but struggle with durable growth
• Allows different parts of the organization to operate at appropriate speeds
• Prevents the entire organization from being constrained by the slowest components
• Creates clear expectations about pace and process for different initiatives

Implementation Requirements

• People in the fast thinking group need to be entrepreneurial and operate with autonomy
• Team members must think holistically about both technical constraints and user experience
• Leaders need to be comfortable with ambiguity and open-ended exploration
• Success requires breaking down role silos and encouraging cross-functional capabilities
• The model works best when paired with a "play with AI" culture where experimentation is encouraged

Measuring Success

• Fast thinking teams: Are they shipping capabilities that create genuine excitement?
• Slow thinking teams: Are they building infrastructure that enables scale and durability?
• Overall: Is the organization becoming more "AI-native" in its approach and execution?

This model represents a fundamental shift from traditional organizational structures that optimize for efficiency within domains to one that balances speed and stability across the entire product lifecycle.