Why per-period attendance beats a morning headcount

Almost every school we onboard starts in the same place: one attendance mark per student per day, taken in homeroom, before the day has really begun. It is fast, it is familiar, and it answers exactly one question — did this child enter the building?

That question matters for safeguarding. It is close to useless for everything else.

What a daily mark hides

Consider two students, both marked present on the same Tuesday. One sat through six periods. The other was in registration, then disappeared after break and missed maths, physics and the lesson where the coursework brief was handed out.

Your register says the same thing about both of them. The maths teacher knows the truth, but that knowledge lives in their head, or in a personal notebook, and it never reaches the person who calls home.

Per-period capture — a session record for each scheduled class, with an absence record against it — changes the shape of what you can see:

• Subject-specific avoidance becomes visible. A student at 94% overall can be at 61% in one subject, and that pattern is a much better early warning than the aggregate.
• Post-break drift shows up as a time-of-day curve rather than an anecdote.
• Teacher-level inconsistency becomes measurable — if one class has near-zero recorded absence all year, that is usually a data problem, not a discipline success.

The objection: "teachers won't do it six times a day"

They will, if it costs seconds. The failure mode is never the concept, it is the interface. If taking a register means loading a page, choosing a class, choosing a date, and ticking thirty boxes that all default to unticked, teachers will do it once a day at best and backfill the rest on Friday.

The rules we hold ourselves to:

• The teacher's class list opens on the current period by default. No date picker, no class picker, no navigation.
• Everyone defaults to present. The teacher marks the exceptions, which on a normal day is two or three taps.
• A late arrival is one tap, not a separate workflow.
• The record closes itself. There is no "submit" that a teacher can forget.

When it takes under fifteen seconds, six-times-a-day stops being a burden and starts being a habit.

The rollups are the actual product

Per-period marks are raw material. What an administrator needs is the derived view: a daily status per student computed from the sessions, and a term-level percentage computed from the days. That way the safeguarding question ("was this child in school today?") and the academic question ("is this child being taught?") are answered from one dataset instead of two competing ones.

It also means the number you put in front of a parent is defensible. "Amir has missed nine of forty-two maths periods this term" is a conversation. "Amir's attendance is a bit low" is not.

Where schools go wrong first

The most common mistake is switching on per-period capture without fixing the timetable underneath it. If the schedule in the system does not match the schedule on the wall, every session record is attached to the wrong slot, and within a fortnight nobody trusts the data.

Get the timetable right first. Attendance is downstream of it, and so is everything else.