Between-session check-ins that sound like you
Behavior change happens between appointments, and consistent, supportive contact is what keeps clients engaged. AI helps draft check-in sequences and non-judgmental reframes at scale. But food and body messages are sensitive: disordered-eating signals and clinical concerns are not something to automate — they go to a human, and often to a mental-health professional.
You write client messages for a nutrition practice. Draft two message sets for this segment: {{segment}}. Match this voice: {{voice}}. These send through {{channel}}. Set 1 — routine check-ins (3 messages): a start-of-week nudge, a mid-week encouragement, and an end-of-week reflection prompt. Focus on habits, consistency, and how the client is feeling about the process — not weight, calories, or any diagnosis. Set 2 — a warm re-engagement message for a client who has gone quiet, with no guilt and an easy way to restart. Rules: - Use [FIRST_NAME] merge fields. Never write a real name or any health detail. - Weight-neutral, non-judgmental language; no shame, no fear-based framing about food. - If a client raises disordered-eating signs or a medical symptom, do NOT counsel it in a message. Draft a version that gently acknowledges it and routes them to me and, where appropriate, a physician or therapist. - One clear action per message. Add "Reply STOP to opt out" to the first message if it is an automated text.
Fill in your details and the prompt updates live — then copy.
Check-in, Monday: Hi [FIRST_NAME], fresh week. What's one meal you'd like to feel good about this week? Reply and we'll make it easy. Reply STOP to opt out. Check-in, midweek: [FIRST_NAME], how are the meals feeling so far — any that clicked? No wrong answers. Re-engagement: Hey [FIRST_NAME], it's been a little while and that's completely okay. Want to pick one small thing to restart with this week? I'm here when you're ready. If a client mentions restricting or purging: "Thank you for trusting me with that, [FIRST_NAME]. I'd like us to talk this through together, and I may loop in a therapist who specializes in this. Can we set up a call?"
The full workflow
- Segment clients in your practice platform (new, engaged, at-risk, lapsed)
- Generate a sequence per segment and edit it to sound like you, not a template
- Load messages with merge fields into your platform — never paste a client list into a chatbot
- Keep the escalation version ready so any disordered-eating or medical signal routes to a human fast
Watch out for
Privacy: no names, diagnoses, weights, or health specifics in a message a bystander could read, and never paste your client list into a consumer chatbot. Generate templates with merge fields and let your platform do the merge.
Don't automate the sensitive stuff: disordered eating and medical symptoms need a human and often a mental-health referral. AI drafts routine encouragement, not therapeutic or crisis communication.
TCPA: automated texts need prior consent and a working opt-out; AI writes the copy but your platform has to honor STOP.
Where this comes from
Every use case on this site is grounded in real reports from working nutritionists — not invented by us.