The quiet cost of tribal knowledge
Public research on knowledge work consistently underestimates how much time is spent navigating around what is undocumented.
Most studies on knowledge worker productivity ask about time spent on specific activities — email, meetings, deep work. Fewer ask the question that matters most for institutional health — how much time is spent reverse-engineering what someone else already knows.
When we conducted our own diary study with 200 knowledge workers across five companies, the median response was 28% of the week. Not 28% of the day searching for files. 28% of the week — across meetings, Slack threads, manager pings, and quiet desk research — reconstructing context that already lived in a colleague's head.
The cost compounds. A junior engineer who spends a third of their week reconstructing context is also depriving a senior engineer of the time they need to do the work the junior cannot yet do. The drag is not just on the individual. It is on the whole team. The remedy is not more documents. The remedy is a captured memory that the junior can actually consult — and that the senior never has to maintain.