Magic & Science Decision Model
by Karri Saarinen • Co-founder & CEO at Linear
Karri Saarinen is the co-founder and CEO of Linear, the cult-favorite issue tracking tool. Previously, he was the founding designer at Coinbase and a principal designer at Airbnb. He is known for his philosophy on 'opinionated software' and design craft.
🎙️ Episode Context
Karri Saarinen discusses how Linear defies conventional startup wisdom by operating with almost no product managers, ignoring A/B testing, and avoiding metrics-based goals. He details the 'Linear Method' of building opinionated software, their unique 'work trial' hiring process, and how they utilized a year-long private beta with cohort-based onboarding to achieve product-market fit.
Problem It Solves
Analysis paralysis caused by over-reliance on data and A/B testing for every feature decision.
Framework Overview
Reject the notion that data must validate every decision. 'Science' is the rigorous gathering of qualitative user insights and understanding the problem. 'Magic' is the leap of intuition taken to solve it without waiting for metric validation.
🧠 Framework Structure
Science is empathy: Everyone in the c...
Magic is intuition: Use the built-up ...
No metrics-based feature goals: Measu...
When to Use
When building 0-to-1 products or when brand/quality is your primary competitive moat.
Common Mistakes
Using data as a crutch to avoid making hard product decisions or taking risks.
Real World Example
Linear built their 'Project Updates' feature based on the intuition that status reporting is broken, without running tests to see if it increased 'daily active usage'.
Sometimes people use data... because they're afraid... 'will I make the wrong choice?' and using data to make the choice for them.
— Karri Saarinen