A certain untransferable, unquantifiable knowledge builds up with using a system repeatedly. 4.2 seconds delay in loading the dashboard. Failure at 47MB when there is a 50MB limit. Missing records are actually in the bin folder after the limit. The comment field has no character count. The third search result always loads more slowly. Clicking “save” twice within five seconds creates a duplicate.

This knowledge isn’t written down anywhere, yet it becomes operational intelligence.

People build workarounds for these. Understanding of delays is especially useful. A mental queue of filler tasks is developed. Clearing notifications = three seconds, reviewing a calendar item = ten seconds, reading a Slack message = twenty seconds.

There are delays that allow switching away (thirty seconds) versus those that demand focus (eight seconds—the system might prompt). The experts initiate delays strategically, starting the report export not when they need it but when they’re about to begin a task that takes two minutes. They batch operations to align processing delays with bathroom breaks.

These users aren’t tolerating the system; they’re choreographing an entire secondary workflow around its delays.