Why 'ghost attendance' numbers from the US don't apply to SA — and what matters instead.

The phrase "ghost attendance" circulates in vendor decks. Its origin is US higher-education financial-aid fraud — the US Department of Education flagged roughly $150M in 2025 aid disbursed against ineligible enrolments. That is a specific fraud pattern in a specific funding system. It does not map cleanly onto what a South African lecturer sees in an NC(V) or NATED class.

The problem DHET actually publishes — and the one procurement teams should be asked about — is throughput. The LMI-Research cohort analysis of the 2016 NC(V) Level 2 intake tracked 88 771 students, of whom 8 135 completed Level 4 in expected time by 2018. That is 9.2% throughput against a 75% National Development Plan target for 2030. You cannot measure throughput honestly if you cannot measure attendance honestly; a register that overstates presence overstates everything downstream.

The product's answer is rotating QR plus lecturer override. Not because we have quantified a ghost-attendance reduction — we have not, and we are not willing to pretend we have — but because those two features together make an attendance record defensible at a DHET audit. A QR that rotates every 30 seconds is uninteresting to screenshot; an override button protects the student whose phone is flat. That is the claim we are willing to make on the record. Everything else the portal does — material, reminders, announcements — sits on top of an attendance record you can actually stand behind.