Designing workflows that survive Term 4

Term 4 is the operational stress test of every school year. The exam coordination, the year 12 wind-up, the yearly reporting, the end-of-year evening events, the stocktakes, the staffing decisions for next year, all compressed into the period when staff are at their most tired.

Workflows that look fine in Term 1 and 2 routinely fall over in Term 4. The interesting question is why.

The pattern I see most often is that the original workflow was designed for a good day. The person designing it was rested, had context, knew the people involved, and could afford to think carefully. The Term 4 version of that same person has none of those things. The workflow that worked at full attention quietly fails at three-quarters attention, because it relied on a kind of vigilance that Term 4 doesn't have.

There are three design choices that make the difference.

The first one. Default to the safe outcome on missing input. A consent form workflow should default to "did not consent" if the form is missing, not to "consented because we didn't hear back". A placement workflow should default to "not allocated" if a host hasn't confirmed, not to "allocated because the email didn't bounce". The cost of being too cautious is one phone call. The cost of being too generous is a child in the wrong place.

The second one. Make the next step impossible to misread. A status of "in progress" tells the next person nothing. A status of "waiting on parent signature, sent 5 May" tells them exactly what to do. The discipline is to encode the action, not the state, in the language. This is not about UX polish. It is about reducing the cognitive load of someone working at three-quarters attention.

The third one. Bake in the handover assumption. By Term 4, the person who started the workflow may not be the person ending it. Maternity leave, illness, retirement transitions, the staff member who moves to a new school in the last fortnight. A workflow designed under the assumption that the same person will run it from start to finish will fall over the first time that assumption is violated.

The cheapest way to test all three of these in your own school's processes is to run a fictional handover exercise. Pick a workflow. Write down what would happen if the owner went on stress leave the day before its busiest moment. If your answer involves "we'd ask them anyway" or "the email trail would explain it", the workflow has not survived its own design test.

Term 4 is too late to fix this. The work is done in Term 1, when there is bandwidth to think about edge cases. The schools that have the calmest Term 4s are not the ones with the most heroic staff. They are the ones whose Term 1 design choices made the heroism unnecessary.