Discovery & Validation Frameworks

Methods to figure out what to build and whether it's worth building before sinking engineering effort.

• Jobs To Be Done (JTBD): Users "hire" a product to make progress in a specific circumstance. Focus on the job, not the demographic. Template: "When ___, I want to ___, so I can ___."
• Continuous Discovery / Opportunity Solution Tree (Teresa Torres): Outcome → opportunities → solutions → experiments, mapped as a tree. Weekly customer contact, not quarterly research.
• Lean UX / Build-Measure-Learn: Hypothesis-driven design; ship the smallest testable thing, measure, iterate.
• Design Sprint (GV 5-day): Map → Sketch → Decide → Prototype → Test. Compress months of debate into a week with a validated prototype.
• Working Backwards (Amazon): Write the press release and FAQ before building. If you can't write a compelling PR, don't build it.
• Story Mapping: Arrange user activities horizontally (journey) and details vertically (priority) to slice releases.
• Crazy 8s: 8 sketches in 8 minutes — forces divergent ideation past the obvious first idea.

How to apply

• Frame the problem as a JTBD statement before designing a solution.
• Use the Opportunity Solution Tree to keep solutions tied to a measurable outcome, not feature requests.
• Working Backwards is the cheapest kill switch: no compelling press release → don't build.
• Story Mapping turns a backlog into a thin end-to-end first slice (pairs with ship-one-vertical-slice discipline).