RFID cards: fast at a gate, useless in a classroom
A card tapped at an entrance gives you a reliable in-time and out-time for the campus. If your actual requirement is "tell a parent their child arrived safely", this is genuinely the right tool.
It falls apart as classroom attendance. Cards get forgotten, lent, and tapped in batches by one helpful student. And campus entry tells you nothing about whether a child attended fourth-period chemistry — which is the number that matters for a board shortfall query.
Biometrics: accurate, and the highest-consequence data you can hold
Fingerprint or face recognition solves the proxy problem that defeats cards. It's accurate and hard to game.
The costs are real, though. Throughput at a single reader is slow enough to matter when 900 students arrive in fifteen minutes, so you need several. Devices need maintenance. And under a data protection regime that treats children's data as a special category, you have chosen to collect and store the most sensitive category of personal data that exists about a minor — one that, unlike a password, can never be reissued if it leaks. That is a defensible choice, but it should be a deliberate one.
In-app marking: only works if it's genuinely fast
The teacher taking the class marks the section on their phone. No hardware, per-period granularity, and the register reflects who was actually in the room.
Its weakness is honest and worth stating: it depends on a human doing it. If marking a section takes three minutes of loading screens, it gets done on Friday from memory and the data becomes fiction. The entire viability of this approach rests on the interaction taking seconds, which is why it's the thing to time during a demo rather than take on trust.
SkoolBook uses this approach. We do not currently integrate RFID or biometric devices — if your school has already invested in that hardware, that's a genuine gap worth raising early rather than discovering later.
Most schools want two things, not one
The requirement is usually two distinct questions wearing one label: "did the child arrive on campus safely" (a parent's question, answered at a gate) and "did the child attend this period" (an academic question, answered in a classroom).
One mechanism rarely answers both well. Schools that recognise this early buy a gate solution for the first and use in-app marking for the second, and are far happier than schools that bought one expensive system expecting it to cover everything.
Separate "arrived on campus" from "attended the period" before you evaluate anything. They're different questions, and one device rarely answers both well.