drg. Rafi & Partners in Medan handles complex cases: jaw correction that combines braces with orthognathic surgery. A single case involves an orthodontist, an oral surgeon, and a hospital, with a treatment journey of 18–24 months. Marketing sent hundreds of leads, the front desk complained they were poor quality, and nobody shared a definition of a “good lead.” We wrote the definition, built a score in the CRM, and taught the ads to find leads that qualify.
The resultQualified leads rose from 23% to 71%, and orthognathic consultations grew 2.4× per month.
Meet drg. Rafi & Partners.
drg. Andreas Tan is managing partner of the group founded by drg. Rafi. The group has two locations in Medan and partners with two hospitals for surgical procedures. Orthognathic patients are usually 18–35 years old with a jaw that protrudes, recedes, or is asymmetrical.
“Jaw correction” ads attracted lots of people who really just wanted regular braces, or were looking for facial plastic surgery. The front desk spent its time explaining that their case wasn't orthognathic. Marketing felt it had delivered leads; the front desk felt the leads were junk. Every monthly meeting ended without a decision.
Marketing counted chats; the front desk counted patients who were a fit.
77% of leads turned out to be regular braces cases or not dental at all.
Google and Meta had no idea which leads were a fit for orthognathic treatment.
Qualified leads were sometimes contacted only 3 days later.
One definition, one score, one weekly meeting.
You can't fix lead quality while nobody agrees on what it means. We started with a written definition, then built the system around it.
- 01Written lead definition
Age, jaw complaint, willingness to share a facial profile photo, and readiness for long-term treatment.
- 02Score in the CRM
The front desk scores leads A, B, or C within 24 hours based on 4 standard questions.
- 03Quality signals sent to ads
Only A and B leads are sent back to Google and Meta as conversions.
- 04Weekly lead review
A 30-minute meeting every Monday between the front desk and marketing, reviewing 10 random leads.
What didn’t work first.
Not everything we tried worked right away. Here are two things we stopped, and what we learned from them.
We added 8 screening questions to the form. Leads dropped 80%, and the ones left weren't any better. We went back to 3 questions and let the front desk ask the rest in chat.
Rejected by Meta under its body-change policy. Replaced with a video of the dentist explaining the signs you need jaw correction.
The engagement timeline.
Lead definition workshop with dentists and the front desk. A/B/C scoring set up in the CRM.
Weekly lead review started. A/B leads sent to Google and Meta.
Ads optimized for qualified leads. Budget moved to the keywords and creatives with the best scores.



