Organizational Memory: Context vs. Knowledge
Teams talk about "losing knowledge" when someone leaves. Usually that's not what happened. They lost context — and mistook it for knowledge.
Two different things
Context is situational and perishable. It's why this decision was made for this customer at this moment. It has a short half-life and it's expensive to reconstruct.
Knowledge is durable and transferable. It's the pattern you can apply to the next situation. It compounds — every project should make the next one cheaper.
Most delivery organizations are drowning in the loss of the first while barely capturing the second.
The disappearing act
Context vanishes because it lives in the seams:
- The hallway conversation that reframed the requirement
- The reason a "simple" change was actually risky
- The tradeoff that everyone agreed to and no one wrote down
By the next sprint, it's gone — and the team re-derives it, badly.
Making memory a system
The goal isn't to write more documents. It's to:
- Capture context at the moment it exists, with as little friction as possible.
- Distill knowledge from context — turn "what we did" into "what we learned."
- Resurface both on demand, so the team decides from memory instead of from scratch.
AI is unusually good at exactly this: reading the raw stream of work and turning it into something the organization can actually reuse.
Context tells you what happened. Knowledge tells you what to do next time. A delivery organization needs both — and a way to move from one to the other.