Every morning at Smile Bright Dental Clinic in Westlands, Nairobi, the receptionist Mary would sit down with her phone and face the same mountain: 40 to 60 WhatsApp messages, most of them sent overnight. Appointment requests, pricing questions, patients asking how to get to the clinic, follow-ups about whether their X-rays were ready. She was often still working through them at 10 AM, by which time new ones had already come in.
Dr. Mwangi, the clinic's owner, put it simply: 'Before Zenai, my receptionist spent her whole morning on WhatsApp before seeing a single patient. We were paying her to be a messaging answering machine rather than front-desk support for real patients in the clinic.'
The Problem in Detail
The clinic was spending roughly 3 hours per day on WhatsApp management. Breaking it down, the messages fell into clear categories:
- Appointment booking requests: 35% of messages
- Pricing questions (consultation fees, procedures): 28% of messages
- Directions to the clinic: 18% of messages
- Follow-ups on test results and X-rays: 12% of messages
- Everything else (cancellations, special requests): 7% of messages
The kicker: 93% of messages were in the first four categories — all of them answerable with the same 4–5 responses, repeated 20 to 30 times each day.
The Setup: Under 1 Hour
Dr. Mwangi's team connected Zenai to their existing WhatsApp Business number and went through the onboarding flow. Within 45 minutes, they had created intent scripts for every high-volume message category:
- Pricing intent: Covers "how much", "consultation fee", "cost of filling", "scaling price" — replies with a formatted pricing table and a booking CTA
- Directions intent: Covers "where are you", "location", "how to get there", "nearest matatu" — replies with a Google Maps link and landmarks
- Hours intent: Covers "open", "closed", "Saturday", "Sunday", "holiday" — replies with full schedule including public holiday notice
- Booking intent: Detects "appointment", "book", "schedule", "I want to come" — triggers the booking flow and logs the request to the calendar
For complex queries — like a patient asking whether a specific procedure was covered by NHIF — the AI fallback kicks in. It reads the clinic's knowledge base (which Dr. Mwangi filled out during setup) and generates a precise, helpful response.
Results After 30 Days
80% of incoming messages are now handled automatically, without Mary touching her phone. She now reviews and responds to the remaining 20% — genuinely complex queries and escalations — in about 15 minutes each morning.
Three additional outcomes surprised the team:
- Appointment no-show rate dropped by 34% — automated reminders went out 24 hours before each booking, something that was previously done manually and inconsistently
- After-hours bookings increased by 60% — patients could now book at 10 PM and get an immediate confirmation, instead of waiting until morning and sometimes forgetting to follow up
- Response time improved from an average of 2.4 hours to under 30 seconds for all automatable queries
What They Would Do Differently
Dr. Mwangi says the only thing he wishes he had done differently was set up Zenai sooner. 'We thought it would be complicated — something that needed IT support or a developer. It took less than an hour and we saw results the same day. The free trial made it risk-free to just try it.'
The clinic now uses Zenai for their second branch in Karen as well, managing both from a single dashboard.