Growth campaigns
Library templates, CRM filters, email + WhatsApp, wallet preview. Best when you want a fixed drip.
Overdue lists: sends and reads, then a real booked slot
Preventive programs re-engage overdue lists with campaigns or autonomous sends; success is sends and reads until a human or workflow marks the appointment booked. The qualifier still collects reason, service_need, location, and preferred_times when they reply. Calendars remain slot truth. Emergencies: call emergency services immediately.
- CRM segments + wallet preview
- Campaigns or autonomous lists
- Booked ≠ send success
Excel recall is not a program, and “booked” is not a send
Hospitals and clinics export overdue annual-check names to Excel and blast from a personal WhatsApp. Vendors then claim autonomous “books the slot.” CoPulse splits the systems: Growth campaigns or autonomous outbound hit the list (wallet preview, Gupshup templates); replies hit the qualifier and appointment skill; Appointment booked is a calendar fact. The bot does not diagnose why they are overdue.
Omnichannel journey
Segment in CRM; WhatsApp/email outbound; qualifier on reply; voice for non-readers; inbox for package questions.
- Segment Recurring patients or a CSV you imported — last consult older than your rule. Preview Organization wallet.
- Campaign or autonomous send Approved template. Autonomous success is sends, reads, or contacted — not booked.
- Reply qualification Conversation agent collects preferred_times and package service_need, then offers calendars.
- Voice for the silent Optional dialer on the same list. Speech-to-text replied ≠ they will attend.
How it runs in CoPulse
Campaigns and autonomous are the outreach runtimes. The conversation agent is the reply runtime. Workflows mark booked when the calendar is true. Do not collapse them.
Autonomous agents
List or objective outbound after you approve a plan. Measure sends and reads. Not “stop when booked.”
Conversation agent
When they reply, qualifier + package catalog + appointment skill + emergency guardrail.
Workflows
On Appointment booked, stop treating them as overdue in your own flags — autonomous will not infer that for you today.
Operator setup
Same Organization as the clinic or hospital Project. Do not invent a second marketing tenant — use a Project or a campaign segment.
Setup → Channels (WhatsApp via Gupshup, email, phone, web, Meta if you run ads) → Library templates → Agents → Studio (qualifier + catalogs + appointment skill + emergency guardrail) → Agents → Router → Setup → Customers → Pipeline → Appointments/calendars → Growth campaigns with Organization wallet preview.
1,200 overdue names, honest meters
You filter Recurring patient with last consult > 12 months, preview wallet, and either start a Growth campaign or approve an autonomous plan on a library template. Day two: 400 reads. That is success for autonomous. Forty people reply; the qualifier books Saturday packages on calendars and the board shows Appointment booked for those forty. The other 360 reads are not booked — do not report them as appointments. A human handles “does this include ECG” in the inbox. No live SMS. No HIS sync claimed.
Keep package names in Services. Do not let the outbound copy promise a clinical outcome. If you also run OPD, router so program replies do not hit the dental agent.
Patient Inquiry Qualifier fields are reason, service_need, location, and preferred_times. The agent does not invent a clinical finding. Knowledge bases hold Departments, Doctors, Beds, and Services catalogs — not a clinical chart. Appointments and calendars are the slot truth; the Appointment booked column is not inferred from a campaign send. The omnichannel inbox keeps WhatsApp, voice, email, and web on one contact so a nurse or desk can take human handoff with the thread. Pricing is wallet-only — no seat fees. Autonomous tasks, when used, count sends, reads, or contacted until a human or workflow moves the pipeline. If a visitor describes an emergency, the agent must tell them to call emergency services immediately; CoPulse is not an emergency line and not an HIS or EMR. Playbook: annual checkup reactivation. Product surfaces: campaigns and autonomous. Never advertise autonomous success as appointment booked. Excel remains a bad source of truth once the Project CRM holds the list. For health-check programs, keep one Organization and a Project per facility or program cluster. Preview wallet cost before any blast. Library templates are required outside the 24-hour WhatsApp window. Workflows fire on stage change only when you attach them under Agents → Workflows. The agent router keeps a specialty bot off a billing thread. Funnel percentages on library kits are estimates, not an SLA.
Program outreach hops
Outbound meters and booking meters are different columns in your head — and in the product.
| Hop | Channel | What happens | Owner |
|---|---|---|---|
| 1 | CRM segment | Overdue list | Operator |
| 2 | WhatsApp / email | Template send | Campaign or autonomous |
| 3 | WhatsApp reply | Qualifier + slot | Conversation agent |
| 4 | Phone | Non-readers | Voice (optional) |
| 5 | Calendars | Appointment booked | Skill / human / workflow |
Playbooks for this team
Jobs are omnichannel. Each link is one URL, not a WhatsApp-only clone.
Guide: Autonomous reactivation
How-to for this desk. Playbooks remain the jobs.
Guide →Annual checkup reactivation
Overdue list outreach with honest send/read success.
Open playbook →Appointment booking
What happens after they reply, not what the blast claimed.
Open playbook →Patient qualification
Qualifier fields on the reply thread after the template send.
Open playbook →Frequently asked questions
Programs, autonomous meters, and wallet blasts.
Does autonomous stop when they book?
No. Success is sends, reads, or contacted. Mark Appointment booked from calendars, a human, or a workflow.
Campaign vs autonomous?
Campaigns are scheduled drips you design. Autonomous is list/objective outbound you approve. Replies still need the conversation agent.
Can we blast SMS to the overdue list?
Connect SMS under Setup → Channels (MSG91), or use Gupshup WhatsApp templates and email.
Is this a diagnostic screening bot?
No. It re-engages a list and books a package slot. Findings stay with the clinician after Consult done.
Who pays?
Organization wallet. Preview before the 1,200-name send. No seat fees.
Where do I configure this?
Growth campaigns or Agents → Autonomous, then Studio for the reply agent, then Appointments/calendars.
Related
Parent hub, sibling teams, and product surfaces this ICP actually uses.
Healthcare hub
Appointments, qualifier, catalogs, and inbox — directory of care-team ICPs and jobs.
Back to hub →Clinics
Day-to-day OPD on the same Project as the program list.
Open →Independent practitioners
Smaller lists, same wallet-only meter.
Open →Autonomous agents
List outbound with send/read success — not booked.
Explore →Wallet-only pricing
Program blasts debit the Organization wallet after preview.
See pricing →Run the overdue list with honest success metrics
Preview wallet, send the template, and book only when the calendar is true.
Get started free