How a Dubai clinic increased bookings 34% with automated follow-ups

Note: The clinic in this case study has been anonymised at their request. We've referred to them as "Bloom Dental" throughout.

Bloom Dental is a private dental clinic in Dubai Marina. When they came to us in November 2025, they had a specific problem: a large gap between enquiries and bookings.

"We get maybe 40 to 50 WhatsApp enquiries a week," the clinic manager told us in our first call. "But we're only converting about 25% of them into actual appointments. The rest just... go quiet."

That's a conversion rate of one in four. With a typical first appointment value of AED 600–1,200, the uncaptured revenue was substantial.

Diagnosing the drop-off

Before recommending anything, we audited 60 days of their WhatsApp conversation history. The pattern was clear:

Most enquiries followed the same arc: a customer would ask about a treatment (teeth whitening, Invisalign, a checkup). The clinic would reply with information and pricing. The customer would say "let me think about it" or simply stop responding. And then nothing — no follow-up from the clinic, no booking from the customer.

The average time between the last message and the conversation going cold: 18 hours. Nobody at the clinic was following up because it felt awkward, time-consuming, and most of the time there was no clear signal that the customer was still interested.

But the data told a different story. Of the 60-day sample, 38% of customers who went quiet had visited the clinic's website again within 72 hours of their last message. They weren't gone — they were still considering. They just needed a nudge.

What we set up

We connected Bloom Dental's WhatsApp Business API to RheXa and configured three things:

1. Immediate AI response layer
Every inbound enquiry now gets a reply within 30 seconds — day, night, weekend, public holiday. The AI is trained on Bloom's full service catalogue, pricing in AED, clinic location, and the team's qualifications. It answers most first-contact questions accurately without any human involvement.

2. Lead scoring
RheXa classifies each conversation by intent level. "What is Invisalign?" is low intent. "How much does Invisalign cost and how many appointments does it take?" is medium-high. "I've been thinking about it for a while, can I come in this week?" is conversion-ready. High and medium-high intent conversations are flagged to the clinic's front desk team in real time.

3. Follow-up sequences
For conversations where the customer went quiet after the initial information exchange, RheXa sends a follow-up at 24 hours. Not a generic "just checking in" — a specific message referencing what they asked about. "Hi [name], wanted to follow up on your question about teeth whitening. We have availability this week on Tuesday and Thursday afternoon — would either of those work for you?"

If there's no response after the first follow-up, a second one goes at 72 hours: softer, no pressure, just re-opening the door. After that, the sequence stops. We don't spam.

The 90-day results

We ran this configuration from December 2025 through February 2026 and compared to the same period the previous year.

Metric	Before RheXa	After RheXa	Change
Weekly enquiries	44 avg	51 avg	+16% (busier season)
Booking conversion rate	24%	32%	+34%
Average response time	4.2 hours	28 seconds	−99%
Follow-up rate	11%	100%	All cold leads followed up
Bookings from follow-up messages	0	31 over 90 days	New channel

The 34% increase in conversion rate translated to approximately 11 additional bookings per week. At an average value of AED 800 per first appointment, that's roughly AED 8,800 per week in revenue that previously went nowhere.

What surprised the clinic

The clinic manager flagged two things that weren't in the original brief:

Saturday afternoons became productive: Previously, Saturday afternoon enquiries sat unanswered until Monday. With the AI handling first contact, Saturday became one of their strongest booking days. Three of the 31 follow-up bookings came from enquiries received on Saturday evenings.

The follow-up tone mattered enormously: In the first two weeks, we tested two versions of the follow-up message. Version A was neutral: "Hi, just wanted to check if you had any more questions." Version B referenced the specific treatment: "Hi, wanted to follow up on your question about Invisalign — are you still considering it?" Version B converted at 3.4× the rate of Version A. Specificity isn't just polite — it's demonstrably more effective.

What the clinic staff actually does now

This is the question we always get: "Did you replace their staff?" No. The front desk team of two people still handles all actual bookings. What changed is what they spend their time on.

Previously: answering the same FAQ questions over and over, missing messages during busy periods, forgetting to follow up.

Now: they focus on the conversations RheXa flags as high-intent, handle any complex situations the AI can't resolve, and manage actual appointment scheduling. The volume of "what is Invisalign?" messages they personally handle dropped to nearly zero.

The AI handles the volume. The humans handle the judgment calls. That division of labour is what made the 34% improvement possible without adding headcount.