Ask a placement cell what they need to know and they will tell you plainly: which of these students can I put in front of a recruiter next month without embarrassing the institution. It is a specific, answerable question. Almost no learning platform answers it.
What platforms report instead is completion. Ninety-one percent of the cohort finished the module. This is a measure of how many people reached the end of a sequence, which correlates with persistence and with how easy the sequence was, and almost not at all with capability.
The substitution happens because completion is trivial to instrument and readiness is not. Completion is a boolean on a row. Readiness requires deciding what capability means for a given role, finding proxies you can actually measure, weighting them, and then defending the weighting when someone disagrees with the result.
It is worth doing anyway. A readiness score that is sixty percent right is far more useful to a placement cell than a completion percentage that is perfectly accurate about the wrong thing.
The version we settled on has four components: reviewed project work, assessment history, consistency over time, and communication. Project work dominates, because it is the only component where a human has looked at something the student made and formed a judgement. The rest are corrections against it.
None of this is settled. We publish the weighting to institutions and expect argument about it, which is the point — a score nobody can interrogate is a score nobody should trust.
LARE Editorial · 15 Jul 2026