Then she accepted a position at another institution.

Her replacement arrived to find nine studies with nine different organizational structures, each reflecting the departing coordinator's evolving filing preferences over six years. Study folders on the shared drive had names like "CARDIO_stuff_FINAL," "IRB_misc_2024," and "Protocol_NEW_v2_REAL." Physical binder tabs were labeled in shorthand that made sense to someone who had spent six years with those studies but communicated nothing to a newcomer. The replacement coordinator, who was competent and well-trained, could not locate a requested IRB approval letter across three studies during her first monitoring visit. Not because the letters did not exist. Because the filing system was not a system. It was one person's mental map, externalized inconsistently across binders and folders, and it left with the person who created it.

This scenario is not unusual. I have seen some version of it at nearly every site I have visited, and it reveals a fundamental design failure that no amount of individual competence can overcome. The departing coordinator did not fail. The site failed to build infrastructure that transcended any individual. That infrastructure is a filing taxonomy paired with naming conventions -- the subject of this lesson.

Then she accepted a position at another institution.

Her replacement arrived to find nine studies with nine different organizational structures, each reflecting the departing coordinator's evolving filing preferences over six years. Study folders on the shared drive had names like "CARDIO_stuff_FINAL," "IRB_misc_2024," and "Protocol_NEW_v2_REAL." Physical binder tabs were labeled in shorthand that made sense to someone who had spent six years with those studies but communicated nothing to a newcomer. The replacement coordinator, who was competent and well-trained, could not locate a requested IRB approval letter across three studies during her first monitoring visit. Not because the letters did not exist. Because the filing system was not a system. It was one person's mental map, externalized inconsistently across binders and folders, and it left with the person who created it.

This scenario is not unusual. I have seen some version of it at nearly every site I have visited, and it reveals a fundamental design failure that no amount of individual competence can overcome. The departing coordinator did not fail. The site failed to build infrastructure that transcended any individual. That infrastructure is a filing taxonomy paired with naming conventions -- the subject of this lesson.