Whatever Works – Scrum, Kanban, or Nothing?
summary
Argues that big tech often lacks a single companywide process and instead has engineers run projects with a simple rhythm: plan build and ship, then measure and iterate, which often outperforms strict Scrum or Kanban. Recommends three concrete priorities: co-own the product and metrics with product people, rebuild first-class developer tooling and architecture competence, and abolish rituals that add bureaucracy. Claims less imposed structure plus engineering craft and ownership produces faster delivery and higher motivation.
AI-written summary · about
chapters (12)
- 0:00Agile and IT challenges
- 1:28Engineer to agile practitioner
- 3:16Why agile transformations fail
- 5:47Leadership in agile
- 8:41Agile in big tech
- 10:48Firms without methodologies
- 12:32Processes purpose and drawbacks
- 16:53One-size-fits-all flaws
- 17:57Project execution lessons
- 20:46Questioning agile dogmas
- 22:36Whatever works approach
- 23:18Conclusion
related talks
42mHow Doing the Little Things Well Makes the Big Things Possible
34mBuilding a Culture For Developer Productivity
45m#NoFrameworks: Build Your Own Agile, Burn the Playbook
21mAutonomous Teams
32mDeveloper productivity 2.0
51mDon't Kill Agility with Agile Processes
01h 04mIs Agile Still the Answer? Lessons from Real World Agility
33mSoftware development practices for companies that want to make an impact