Patient recall automation on WhatsApp
Why WhatsApp is uniquely suited to recall
SMS reads are perfunctory. Email open rates in India hover at 12-18% for healthcare and half those clicks. WhatsApp utility opens run 65-85%. When the recall is tied to a genuine service event, patients engage because the message is useful.
The catch: WhatsApp punishes over-messaging and promotional-toned recalls harder than any other channel. Quality rating drops from Yellow to Red inside a week if a clinic blasts a marketing template to a poorly-consented list. Recall done well converts; recall done badly kills the number.
What counts as a compliant recall
Utility recalls — tied to a service event, no opt-in gate
Meta and DPDP both accept recalls that reference a specific service moment — a scheduled 30-day post-op check, a 90-day braces adjustment, an annual dental cleaning that was booked as part of a package. These fit the utility category because the patient's active service relationship anchors the message.
Example (utility, passes):
"Hi {{1}}, this is a reminder that your 3-month dental follow-up is due. To schedule, reply BOOK or call {{2}}."
Marketing recalls — need explicit prior opt-in
Recalls that offer a discount, cross-sell a new service, or invite the patient to a camp are marketing category. They need documented DPDP-compliant consent captured at intake with clear purpose disclosure. The consent record must be produceable.
Example (marketing, needs opt-in):
"Hi {{1}}, our monsoon wellness package is live — 20% off full body checkup this month. Reply INFO to know more."
Recall cadence by specialty
The right interval matches the natural cycle of the service. Sending a follow-up recall too early feels intrusive; too late and the patient has already gone to a competitor.
| Specialty | Recall cadence | Category |
|---|---|---|
| Dental — routine | 6-month cleaning + hygiene | Utility |
| Dental — implant | 30-day, 90-day, 6-month, annual | Utility |
| Dermatology — treatment | 3-month follow-up, seasonal skin health | Utility + marketing |
| Aesthetic — injectables | Procedure-linked (4-6 month typical for maintenance) | Utility |
| IVF | Cycle-linked, next-cycle window | Utility |
| Ophthalmology — LASIK | 1-week, 1-month, 6-month, annual | Utility |
| Paediatric | Vaccine schedule (age-based) | Utility |
| Cardiology | 3-month medication review, 6-month lipid panel | Utility |
| Hair transplant | 1-week, 1-month, 3-month, 6-month, 1-year | Utility |
| General preventive | Annual health check | Utility |
The three-touchpoint recall sequence that works
Touch 1 · Day of recall trigger
First message. Warm, service-anchored, one clear action. Utility template, fires from CMS event or scheduler.
Touch 2 · Day +7 if no response
Second message. Slight variation on the message body, references the same recall reason. Fires only if no engagement on the first message.
Touch 3 · Day +21 if no response
Third and final message. Adds a soft doctor-mentioned line ("Dr {{2}} recommends this check to..."). After this, the patient goes back into the passive base — no further active recall for that cycle.
The stop condition matters. Four or more recall messages without engagement damages quality rating and increases opt-out. Three is the ceiling.
Segmentation that lifts response rate
Blasting the entire recall cohort with one message pattern underperforms. Segmentation options that pay for themselves:
- By procedure history — an aligner patient gets aligner recall, not a general dental recall
- By last-visit recency — 6-month lapsed vs 18-month lapsed need different tone
- By channel of original acquisition — Google Ads leads respond better to structured recall; word-of-mouth leads respond better to warm language
- By clinic location for multi-city — one branch's recall should not carry another branch's contact number
How CMS integration changes the game
Manual recall (someone at reception typing patient names into a WhatsApp template) is possible at low volume but breaks past ~200 patients. Automated recall requires clean patient data in the CMS with the right event fields:
- Last visit date
- Procedure type (categorised, not free text)
- Doctor of record
- Marketing opt-in flag (for marketing recalls)
- Preferred language (English, Hindi, regional)
- Do-not-contact flag (respected forever)
The CMS fires "patient.recall.due" events on a scheduler; the BAPI receives the event and triggers the segmented template. On PatientPulse Growth and Scale we wire this up as part of the standard integration.
Measuring recall correctly
- Recall delivery rate — templates delivered vs sent (should be 96%+)
- Recall read rate — read vs delivered (should be 60%+ within 24h)
- Recall response rate — replies vs delivered (varies 5-20% by segment)
- Recall-to-booking conversion — booked appointments vs delivered (target 8-18%)
- Opt-out rate — should stay under 1% per campaign; over 3% is a quality-rating warning
- Quality rating trend — must stay Green
Common recall automation mistakes
- Sending the same generic template to the entire base — patients notice, opt-out rate spikes
- Marketing templates sent to non-opted-in list — DPDP breach and quality rating hit
- No stop condition — repeated messages with no engagement erode quality rating
- Recall message includes a price — often gets reclassified from utility to marketing
- Recall sent at odd hours — 11 PM template send lands in the middle of an angry patient's night
- No reception handoff plan — recall generates enquiries no one is ready to convert
Turn the dormant base into next month's revenue.
PatientPulse Growth includes recall sequence design, CMS wiring, template authoring and monthly measurement. Custom-scoped, with retainers from ₹20,000/month.
See PatientPulse pricing → Chat on WhatsApp