Beyond Features
summary
Claim: software delivery must treat three kinds of work as first-class — features discovery kaizen — and optimise for sustainably minimize lead time to business impact rather than for ritualised predictability. Evidence: Toyota stops the line ~7,000 times/day to improve flow; run parallel experiments (example: three pairs implement the same feature in different JVM languages for 4–6 weeks) to get evidence; use a lower/upper bound estimate (often 2–4 months) and front‑load discovery to reduce overall lead time.
AI-written summary · about
chapters (10)
- 0:00Start
- 2:46The Agile Manifesto
- 5:05Setting Up a Continuous Integration Pipeline
- 7:04Contract Negotiation
- 8:54Culture Eats Strategy for Breakfast
- 10:22Relationship with the Business
- 13:38Predictability
- 22:39Business Impact
- 32:54Software Development Is Generative
- 34:09Complex Adaptive System
related talks
45mThe Goldilocks Problem of Software Delivery
47mContinuous Deprovement will eat your product for breakfast
30mDelivering Features VS Delivering Impact
31mBalancing New Features While Maintaining Quality
26mFeatures Versus Futures
22mOptimizing Product Development: Balancing Speed, Quality, and Process
29mEconomies of Speed
51mImproving Software Flow