Scaling Agile: The Small-is-Beautiful of Hubs
summary
Argues that hierarchical scaling frameworks produce long communication chains — an example shows average 7.6 hops between people in a 625‑person Scrum@Scale hierarchy — and proposes using self‑organizing hubs from scale‑free networks to cut graph diameter and preserve fast feedback. Reports sociometric analyses of ~200 organizations (examples: Borland Quattro Pro, an engineering documentation group, early AT&T XP project), quantifies a ~40% coordination overhead when teams split, offers the Pasternet social network analysis tool (Mac App Store) to model hubs and measure fragility vs robustness.
AI-written summary · about
chapters (14)
- 0:00Intro
- 1:39Scale-free networks
- 6:22The problem
- 15:15Scaling and scoping
- 18:33What are scale-free networks?
- 23:31Hubs reduce graph diameter
- 25:13Node degree
- 29:39Examples of scale-free networks
- 32:22What are they good for?
- 46:33We already know some examples
- 49:22Example: Amazon, Apple, FB, Google, Microsoft & Oracle
- 50:16Do it at home
- 50:55Conclusion
- 52:15Outro
related talks
49mNet Positive Development
48mScale, Microservices and Flow
53mKeynote - Software Architecture, Team Topologies and Complexity Science
49mSoftware Architecture, Team Topologies & Complexity Science
58mScaling Agile in Organizations
42mScale, Flow and Microservices
51mTech-enabled Network of Teams
37mScale, Flow and Microservices