16 September 2026

School Attendance Alerts Without Spamming Parents

An attendance alert system should tell a principal an aggregate number, tell a teacher a specific student and a suggested next step, and tell a parent something supportive about their own child only — never a daily message, never a comparison to another student or a class average, and never language that implies fault.

Attendance is one of the easiest things for a school system to alert on, and one of the easiest to get wrong by over-alerting. A daily "your child was absent" message for every absence quickly becomes background noise a parent stops reading, and a message that compares one child’s attendance to a class average risks reading as an accusation rather than information. The right design is not "alert more" — it is alerting the right person, at the right level of detail, in the right tone.

One fact, three different messages

The same underlying signal — a student’s attendance has fallen below a threshold over a rolling window — should reach three different audiences in three different ways.

• Principal: an aggregate view — "6 students across Class 8 crossed the attendance threshold this week" — to allocate attention, not 40 individual pings.
• Teacher: a specific student and one concrete suggested next step, readable on a phone in a few seconds.
• Parent: a supportive, specific message about their own child only — no class average, no rank, no comparison to any other student.

Why the parent message is the one that needs the most care

An attendance alert is, underneath the framing, a form of monitoring a child’s behaviour. Sent supportively — "we noticed a few missed days this month, is there anything the school can help with" — it opens a door. Sent in a risk-flavoured tone, it produces defensiveness, and in the worst case, a child being punished at home for what was meant to be a helpful notice. That outcome runs directly against the point of sending it at all.

Why this should not go straight to parents by default

Because attendance-based flagging is a form of behavioural monitoring of a child, and monitoring a child’s behaviour touches specific obligations around consent under India’s data protection law, the sensible sequencing is to let staff see and act on the signal first — where there is a clear educational purpose — and keep automatic parent-facing dispatch behind a setting that stays off until the consent basis for it is specifically confirmed. This mirrors the same sequencing used for the Haryana Form 6 fee advisory: staff-facing insight can move faster than anything reaching a parent automatically.

The threshold itself should be a setting, not a constant

A single hardcoded attendance percentage baked into the code assumes every school’s context is identical, which it is not — a school in an area with frequent transport disruption may reasonably want a different threshold than one without that issue. Treating the trigger — the attendance percentage, the window it is measured over — as a per-school configurable value, with a sensible default, avoids forcing every school into the same assumption and avoids a future change requiring a code deployment rather than a settings change.

Guarding against false positives

A threshold applied blindly will catch students whose real story is not attendance at all — a student with very few graded assessments in a window, for instance, can produce a misleading signal. A sensible design excludes students with too little data to say anything meaningful, rather than flagging them anyway with false confidence.

What this connects to

The same non-comparative, supportive tone rule applies to how we handle student recognition — a parent-facing message about their own child, positive or concerning, should never lean on a comparison to someone else’s child to make its point. And the same "aggregate for principals, specific-but-careful for the person closest to it" pattern shows up again in how we handle the Daily Learning Pulse, where honesty about what actually happened in class matters more than volume of messaging.