Signal
A risk score tells you a student is in trouble.
It doesn't tell you why — so nobody does anything.
Signal classifies why a student is disengaging, against Tinto's theory of student departure — then hands the result to a workflow that acts.
Three signals, one barrier. The barrier determines the intervention.
The difference
A score tells you who. Signal tells you why.
A risk score
A number. A ranked list. A dashboard someone has to remember to open.
It tells you who is in trouble. It doesn't tell you why — and without the why, the number doesn't tell you what to do next.
So the list gets long. The dashboard gets stale. The student is already gone.
Signal
A named barrier — academic, financial, social, or life-event — tied to a specific student, handed to a workflow that fires.
The barrier determines the intervention. The intervention reaches the student over SMS. The student responds.
Detection that names the barrier is a different thing from detection that produces a score.
The four barriers
Disengagement has a shape. Signal names it.
Signal classifies disengagement against Tinto's theory of student departure. Four barriers. Each one determines a different intervention.
Academic
What it looks like
Missed assignments, falling grades, LMS logins gone cold. The student is losing ground in the classroom.
Example
A student who logged in daily stops logging in for nine days after a failed midterm.
What Reach does
Sends an academic check-in with tutoring slot links.
Financial
What it looks like
Aid file incomplete, registration hold, account balance unpaid. The student is being blocked by money.
Example
A Pell grant file is missing one document. A hold is placed. The student doesn't know why they can't register.
What Reach does
Sends a specific message naming the missing document and the link to fix it.
Social / belonging
What it looks like
Attendance dropped, no campus engagement, course load reduced. The student is pulling away from the institution.
Example
A first-year student stops attending a gateway course and drops from 15 to 9 credits.
What Reach does
Sends a belonging check-in and routes to a navigator for a conversation.
Life-event
What it looks like
Sudden disengagement across all signals at once. Something happened outside of school.
Example
A student who was fully engaged stops all activity mid-semester — no logins, no submissions, no attendance.
What Reach does
Flags for human outreach. Routes to a navigator, not an automated message.
What Signal sees
The behavioral inputs.
- LMS activitylogins, assignment submissions, time in course
- Registration stateenrolled, dropped, hold
- Financial aid file statuscomplete, incomplete, pending
- Account holdstype and date placed
- Attendance patternscourse-level, week-over-week
- Credential progresscredits earned vs. attempted
Send a CSV. That's the setup.
Boom starts from a flat export you can produce today. No data warehouse project, no middleware license, no line item in next year's IT roadmap. Most campuses are live in 48 hours.
Start with a CSV. Move to scheduled exports when you're ready. Deepen into direct SIS reads when and if the campus wants to. Boom grows into your stack on your timeline.
No VPN, no firewall changes, no database credentials. If you can email a spreadsheet, you can run Boom.
Evidence
Early results from a pilot community college partner.
~41%
reduction in summer melt compared to the control group, measured in a quasi-experimental design at a pilot partner institution.
Preliminary and in progress
These are preliminary results from an ongoing quasi-experimental study. The sample is a single institution. The study has not been peer-reviewed. We're sharing it because institutional buyers deserve to see the evidence as it stands, not after it's polished. Full methodology is available on request.
Matthew West
Founder · school psychologist, 14 years
I spent fourteen years as a school psychologist on under-resourced campuses in Los Angeles. I watched students disappear. Not drop out — disappear. They'd stop logging in, stop responding, stop showing up. By the time anyone noticed, they were gone. The signs were always there, weeks earlier.
Boom is the detection system I wished existed. It watches the signals. It reaches students over SMS before the drop. And it gives the people on your team the time to do what only humans can do.
Bring us a semester of your data.
We'll show you who you were about to lose.
No commitment. We'll run Signal on your data and show you the results.