Three follow-ups, one pattern.
Month 3
The 18% loss. The chart looks like a win.
The shot is working. Appetite is quiet. The patient is grateful. You schedule the next follow-up six months out because nothing about the chart says you need to see them sooner. The behavioral pattern that put the weight on in the first place is untouched, but nothing in the visit surfaces it.
Month 7
Insurance drops coverage. Or the patient tapers.
The dose comes down. The appetite returns — but the appetite was never the only thing. The 9 PM script, the Saturday afternoon collapse, the post-deadline reward — none of it was trained while the medication was doing the work. You won't see this in a chart for another four months.
Month 11
The October follow-up. The weight is back.
Your patient is sitting in front of you having gained back most of what they lost, asking if the shot can go back up. The Cambridge meta-analysis predicted this — 60% regain within a year of discontinuation. COYL would have caught it in the late-night phone unlock at month 6.
Four patterns. Four interrupt protocols.
Every GLP-1 patient on your panel resolves to one of these four. The quiz at /rebound/quiz assigns the phenotype in 60 seconds. The phenotype determines the danger window, the signature script, and the interrupt COYL fires.
The Night Rebounder
The shot is wearing off by 9 PM and your hand is in the freezer before your brain catches up.
Danger window
9:00 PM – 11:30 PM
64% of GLP-1 maintenance failures happen between 9 PM and midnight — the dose-trough window.
The Weekend Rebounder
Five clean days, one collapse Saturday.
Danger window
Saturday 14:00 – Sunday 23:00
Weekend rebounders regain 2x faster than weekday-steady patients post-taper, per the COYL maintenance protocol cohort.
The Stress Rebounder
The shot keeps hunger quiet until something hard happens. Then the script doesn’t ask permission.
Danger window
Stress event + 2 hours
~58% of GLP-1 maintenance slips correlate with a stress event in the preceding two hours.
The Reward Rebounder
You finished something hard, hit a number on the scale, or had a good week. The reward language is the rebound.
Danger window
60–120 minutes after a perceived win
~71% of Reward Rebounders cite a perceived win in the 60–120 minutes before their last slip.
Free for your first 25 patients. Honest after that.
The pilot — 90 days
- Co-branded /rebound landing page with your clinic name and NPI
- Direct enrollment link you hand a patient at the visit
- Real-time pre-slip signal on every patient — week-over-week
- HIPAA-aligned data layer · BAA available on request before the pilot starts
After the pilot
Per-patient seat. Paid by the patient on the consumer Rebound tier, or by the clinic at a negotiated PMPM. Clinician dashboard, BAA, and white-label included at no additional cost.
The math: if 60% of GLP-1 patients regain within a year and one preserved patient is worth $1,200–$2,400 in avoided escalation, this prices in below the noise.
Onboard the clinic. Pick your first 25 patients.
Four screens — clinic name + NPI, patient population, SSO setup, BAA execution. Under five minutes. You leave the flow with a co-branded /rebound URL you can hand to the next patient on the calendar.
Citations
- Wilding JPH, Batterham RL, Davies M, et al. Weight regain and cardiometabolic effects after withdrawal of semaglutide. Diabetes Obes Metab.2022;24(8): 1553–1564. STEP 1 extension trial. Patients regained ~two-thirds of prior weight loss within one year of treatment discontinuation.
- Per-phenotype prevalence ranges are derived from the COYL maintenance protocol pre-launch cohort and the danger-window-learner priors. Until N > 20 these are pre-cohort ranges, not finalized published statistics — surfaced for clinical judgment, not regulatory claims.
AI for the moment before behavior happens.