Operating model·4 min read
Your architecture will look like your org chart whether you like it or not
Conway wrote this down in 1968. Sixty years on, most target architectures are still drawn without anyone checking who will be allowed to talk to whom.
Source
How Do Committees Invent?
Melvin E. Conway · Datamation · 1968
Read the originalAslanWay is not the author of this work. The summary below describes the source; the commentary that follows is ours and is not endorsed by the author.
What it argues
Melvin Conway's paper is short and its central claim is one sentence. Systems end up mirroring the communication paths of the organisation that builds them — not the paths on the org chart, but the ones people actually use.
“Organizations which design systems (in the broad sense used here) are constrained to produce designs which are copies of the communication structures of these organizations.”
Conway's reasoning is mechanical rather than cultural. Any interface between two subsystems requires the people responsible for them to agree on it. Where that conversation is easy, the interface will be rich; where it is hard, the interface will be thin, rigid, or replaced by duplication.
The paper also makes a second point that gets quoted less: the design is fixed early, when the least is known, and the organisation that produced it is very difficult to change afterwards.
What we do with it
- If the target architecture assumes two teams will collaborate daily and those teams sit in different divisions with different budgets, the architecture is not going to happen. Change the reporting line or change the architecture.
- The corollary is usable: to get a modular system, give the modules to teams that do not need to negotiate constantly. This is the reasoning behind the platform and stream-aligned team split in Team Topologies.
- We test for this in the Frame phase. If a proposed boundary cuts across an organisational fault line, it either gets moved or it gets an explicit owner with authority on both sides.