Attendance Tracking App

To turn daily frustration inside MAAC's attendance system into a defensible brief before designing a single screen.

Was this my irritation, or a structural failure affecting students, staff and management alike?

All three groups were chasing each other for the same data — students chased staff, staff chased management, management chased month-end summaries that arrived too late to act on.

The project stopped being 'redesign the attendance screen' and became 'give one source of truth to three audiences with different jobs'.

Nothing yet — this stage produced the problem statement and the three-audience frame that every later decision is measured against.

To translate 'this is annoying' into hours, delay and trust — the three currencies management actually budgets against.

What does the fragmentation cost per week, and which cost is the one a decision-maker would act on?

~6 admin hours a week lost to reconciliation, data reaching management 3–5 days late, and parents only seeing status at month-end.

The lateness — not the hours — became the headline. Late data turns a preventable conversation into a compliance one.

A cost frame with three lines — time, decision latency, perceived care — that every later feature had to improve at least one of.

To replace my insider assumptions with the actual vocabulary and workarounds of students, staff and management.

What does each audience open the system to find out, and how many steps does that answer currently cost?

The same three complaints in every session — multi-screen navigation, buried history, no trend line. Students need one number; staff need one combined view.

I stopped designing a records browser and started designing a status answer. Browsing is the exception, not the default.

Two grounded personas — Student (needs the % at a glance) and Employee/Staff (needs leave plus attendance in one place) — with their current step counts attached.

To check whether an existing category already solved this, and to find the gap worth designing into.

Do school ERPs, HRIS tools, tap-in apps, biometrics or paper registers optimise for daily use or for reporting?

Every category optimises reporting. None optimises the daily loop — biometrics log perfectly and produce dashboards nobody reads.

Positioning shifted from 'a better attendance app' to 'the daily-use layer the reporting tools never built'.

A positioning statement and a single differentiating metric: admin hours per week reclaimed from reconciliation.

To collapse a five-step hunt into a one-tap answer without hiding anything staff legitimately need.

What is the minimum set of surfaces that serves both personas, and what belongs one level down?

Six surfaces covered every real task; everything else in the old system was a variation on reporting.

Dashboard became the home of status, not records — history and exports moved a deliberate level below.

Six surfaces — Dashboard, Mark Attendance, History, Reports, Notifications, Profile — with everything critical one tap from home.

To move at team pace as a solo designer without outsourcing a single design decision.

Which parts of the process genuinely benefit from acceleration, and which degrade when rushed?

Persona pressure-testing, IA mapping and first-pass copy accelerate well. Touch targets, contrast and tone do not.

I split the workflow explicitly — AI for breadth and speed, manual craft for everything a user physically touches.

A refined Figma system with 44px touch targets, WCAG AA contrast, and copy validated against each persona's real vocabulary.

To make the trade-offs legible instead of presenting the final screens as inevitable.

Which conventional patterns actively hurt this context, even though they look responsible on paper?

Confirmation dialogs, comment fields and dense tables all tested as friction with no recovery benefit on mobile.

Default to the fewest possible taps and let the exceptions live one level down, rather than protecting against rare mistakes.

Dashboard-first over list-first, one-tap marking over multi-step, cards over tables, and exports scoped to Reports.

To confirm the two-tap promise survived contact with people who weren't in the design process.

Can both personas complete their five core tasks unaided, and where do they pause?

All critical flows cleared ≥90% task success at 4.5 / 5 satisfaction; the only hesitation was on export labelling.

Nothing structural — the IA held. Micro-copy on Reports changed, which is the outcome you want at this stage.

Copy revisions on the Reports surface and clearer confirmation feedback after marking attendance.

To commit in advance to the evidence that would show this worked, including evidence that could embarrass me.

Which single metric would a centre head check first, and which one actually reflects student outcomes?

Admin hours are the easiest sell, but time-from-absence-to-intervention is the metric that changes a student's term.

Success got redefined around prevention rather than efficiency — saved hours are the by-product, not the goal.

A four-metric measurement plan with explicit targets and a quarterly before-and-after cadence.