The Challenge
The Segment founders went through Y Combinator with a classroom tool that nobody wanted. They pivoted to a web analytics dashboard — but the analytics market was crowded with Google Analytics, Mixpanel, and Amplitude.
In desperation, they open-sourced their JavaScript analytics library (analytics.js) and put it on Hacker News. The library went viral — but not for the analytics features. Developers loved that it could send data to multiple analytics tools simultaneously with one API.
The Approach — Tools in Action
- "Why are developers downloading analytics.js?"
- For the analytics dashboard? → No, few converted
- For the analytics.js library itself? → Yes, but why?
- For the analytics features? → Somewhat
- For the multi-destination data routing? → Overwhelmingly yes
The insight: developers didn't want another analytics tool. They wanted a way to collect data once and send it everywhere — to any analytics, marketing, or data warehouse tool they used.
Abstraction Laddering clarified the opportunity:- Going up: "Why do developers want data routing?" → "To avoid implementing 10 different SDKs" → "To simplify their data infrastructure" → "To have a single source of truth for customer data"
- Going down: "How do we build a customer data platform?" → "One API that collects events and routes them to 300+ destinations"
The OODA Loop of three pivots was itself the learning process: each pivot (classroom tool → analytics → data routing) was an iteration that got closer to the real problem worth solving.
The Outcome
Segment's third pivot found gold:
- Acquired by Twilio for $3.2 billion in 2020
- 20,000+ companies used Segment to collect and route customer data
- Integrated with 300+ analytics, marketing, and data tools
- Became the standard "customer data platform" (CDP) category
- Proved that the most valuable feature of your product might not be what you think it is — listen to what users actually use
The founders' willingness to pivot (three times!) at YC, combined with the discipline to listen to user behavior rather than their own assumptions, turned a failed classroom tool into a $3.2B acquisition.
Key Takeaway
When users use your product differently than you intended, pay attention — they might be showing you a bigger opportunity. Use Issue Trees to analyze what users actually value, and have the courage to pivot toward it, even if it means abandoning your original vision.
Tools Used in This Story
Issue Trees
Problem SolvingStructure and solve problems in a systematic way
Abstraction Laddering
Problem SolvingFrame your problem better with different levels of abstraction
Confidence determines speed vs. quality
Decision MakingDetermine a trade-off between speed and quality when building products
OODA Loop
Decision MakingMake faster decisions with incomplete data