The way you organize your teams implies requirements on your architectural decisions
Mel Conway published in 1968 the observation that the communication structure of a company has an impact on the architecture of the produced software components.
Nowadays many organizations use the “feature team approach”, that is, a preference for cross-functional teams which deliver value without depending on other teams. An implication is that several feature teams are going to work on the same components, which implies that the traditional concepts of strong component ownership are no longer applicable. This generates a myriad of challenges, e.g. teams waiting for code review from other teams.
In this talk we first revisit Conway’s Law and then explore what is required to ensure success of feature teams working on shared components. First and foremost, does the architecture, the processes, etc. really support the requirement of ease of changing a shared component?