All · Student Support Services
Feedback as an early-warning system
A once-a-year survey tells you who struggled after they’ve gone. Continuous check-ins surface the signals (falling belonging, rising stress, the exam-window wobble) while there is still time for a person to act.
The hardest thing about keeping a student engaged is timing. By the time a student has quietly checked out, the experience that lost them happened weeks earlier, and an annual survey only confirms it after the fact. The case for measuring continuously is not that it produces more data; it is that it moves the signal forward in time, to a point where a person can still do something about it.
The signal arrives weeks before the decision
Three of the patterns elsewhere in this hub are, read together, an early-warning system. Belonging tends to fall before a student disengages, not after. Stress rises into the assessment windows. And exam-readiness confidence has a predictable seasonal collapse, from above 8 at the start of term to the low 6s in December and 5.3 in April. None of these is a surprise on the day it happens; each is visible in the weeks before it bites.
Mean of exam-related check-in questions, 0 to 10. The troughs are the December and April assessment windows, visible in advance, every year.
A team watching those three signals does not need to predict who will leave. It needs to notice, early, the student whose belonging is sliding and whose stress is climbing in the same few weeks, and reach them while that is still a conversation rather than a conclusion.
The system already routes help to the moment
This is not hypothetical. When a check-in answer suggests a student could use support, the platform surfaces a relevant nudge in the moment, and the pattern of where that help is triggered is itself a map of risk: belonging triggers the most of any topic, around 22% of all nudges, ahead of confidence, support services and stress, and the volume peaks in February and around the April exam window, the same two pressure points the confidence data flags. The demand for help shows up exactly where, and when, the experience data says it should.
What it means for institutions
The value of continuous feedback is not the average at year-end; it is the early signal during the year.
- Watch the leading signals, not the year-end average. Belonging, stress, and exam-readiness confidence move weeks before a student disengages, so a team tracking them sees the conditions forming rather than confirming them after the fact.
- Act on co-occurring signals. Falling belonging plus rising stress in the same weeks is a stronger early warning than either alone; reach the student while it is still a conversation rather than a conclusion.
- Treat the help-nudge map as a risk map. Where and when help is triggered tracks the same pressure points the experience data flags, so the nudge pattern points teams to the moments that need attention.
The honest claim is narrow and strong: we don’t predict who will leave, and we don’t measure who does, but we surface the conditions that put a student at risk, falling belonging, rising stress, the exam-window dip, early enough, and specific enough, for a human being to act. For wellbeing and student-success teams, that is the difference between a report and an intervention.
How we measure it
Based on continuous student check-ins on a 0 to 10 scale and the help nudges surfaced from them, over the 2025/26 academic year, mapped to the StudentPulse framework. Figures are shares and aggregates, not raw counts. We describe early signals associated with disengagement risk; the data does not record enrolment or completion outcomes, and associations are not proof of cause. Cross-institution aggregate; no single institution identified.