Our program runs multiple student cohorts in parallel. After every batch finishes, we collect feedback on course content, instructor quality, practical sessions, project support, and materials. We'd been doing this for a while, but what we hadn't been doing was comparing it.
Each batch's feedback sat in its own pile of forms. If you wanted to know whether Batch B was struggling with something Batch A had already solved, the only way to find out was to manually dig through forty-plus individual responses and try to spot a pattern by eye.
Nobody ever actually did this, so the same issues kept surfacing, batch after batch, without anyone connecting the dots.
The module I built
Feedback gets bucketed automatically into five categories: Course Content, Instructor Quality, Practical Sessions, Project Support, Materials.
Cohorts are compared side by side instead of one at a time.
Batch 2024-A finished at 4.5/5 average with 86% satisfaction. Batch 2024-B, still ongoing, is at 4.3/5 and 79% satisfaction—a small gap you'd never notice just skimming individual forms.
-
Why the comparison alone wasn't enough
A chart showing two numbers next to each other isn't useful on its own—you need to be able to ask "did we already know this was coming?" without re-reading old forms by hand.
That's where the shared memory layer mattered: the Batch Improvement Log pulls from the same memory as the Lead Tracker and Student Progress modules (built on Hindsight), so a mentor can ask a question like that and actually get an answer instead of a research project.
Top comments (0)