The owner-doctor said it plainly on our first call: "I've had three agencies in five years. They all post pretty pictures. My schedule is still 60% full and my hygienists are underbooked. I need patients in chairs, not likes."
The dental space has a specific pathology around social media agencies. Most sell a monthly deliverable — 20 Instagram posts, 4 reels, a "reputation dashboard" — and none of it is measured against patients booked. The doctor knows the numbers don't tie out but doesn't have the tools to prove it, so they keep paying the invoice.
What the previous agency was actually delivering
I asked for 90 days of the agency's deliverables and pulled every post's engagement, every DM, and every Google review reply timing. Here's what was actually happening:
Look at those last two lines. 47 people said "hi, I'm interested in becoming your patient" — and 31 of them got no answer for a business day. Three became patients. The agency was doing its content job. Nobody was doing the answer-back job. The pretty pictures were reaching people; the practice couldn't catch what the pictures pulled in.
What we did in 11 days
MyPost Social is the content half. Glovecars is the response half. For this practice we deployed both, since the diagnosis was symmetric — the schedule needed better content pulling in and better response catching.
Day 1-2: Audit + brand voice capture
I spent a Tuesday afternoon in the practice. Recorded 12 minutes of Dr. Patel's front-desk voice for the SMS + DM voice model, walked the office to photograph the actual operatories and hygienists (real people, real chairs — nobody wants stock dental photos), sat with the office manager to pull their PPO/FFS insurance mix and average new-patient production numbers ($720 PPO, $1,850 FFS after first year).
Day 3-5: Content engine + response stack
Set up MyPost Social to publish 4x/week to Instagram, 3x/week to Facebook, 2x/week to TikTok. Content mix: 40% educational (before/after with consent, oral health tips), 30% office culture (hygienists, front desk, dogs — Dr. Patel's golden retriever is a bigger draw than any bleaching promo), 20% patient testimonials (video, filmed in-office with waivers), 10% offers/openings.
Wired Instagram DMs, Facebook Messenger, Google Business Profile messages, and website chat into a unified inbox. The AI (voice: Rachel with a mild upstate NY inflection, since Saratoga patients respond to it) responds within 45 seconds, warm and helpful, discloses AI status, and books directly into their Dentrix system.
Day 6-9: Test + tune
I sent 40 test DMs myself with realistic patient personas (pediatric mom, price-shopper, dental-anxious adult, insurance switcher, new to town) and listened to every AI response. Adjusted three things:
- The AI was too fast to offer specific price ranges — patients experienced it as pushy. We tuned it to acknowledge the price question, offer one exam-only quote ($95 no-insurance new patient), and route the rest to a callback.
- The AI was giving out too much clinical information for the dental-anxious persona. We restricted it to soft acknowledgment ("we hear that a lot — Dr. Patel has a very gentle approach and we have IV sedation on-site") and a "schedule a 15-minute meet-and-greet with the doctor" offer.
- Pediatric moms respond dramatically better when the reply says "we have a small waiting room for the little ones with quiet toys, not iPads" — this line pulled +40% booking rate. Small thing. Big signal.
Day 10-11: Live
Content started publishing Monday. Response stack went live Tuesday afternoon. First same-day DM booking landed 6 minutes after the AI came online — a pediatric mom who'd been price-shopping for two weeks.
What happened in 90 days
New patients: 18/month → 34/month. Of the 34, roughly 15 were FFS (higher margin) versus the previous mix which was ~70% PPO.
- DM-to-consult rate: 6.4% (previous agency era) to 14.2% (with MyPost Social + response stack).
- Google reviews: +47 new 5-star reviews across the 90-day window — a byproduct of the AI's post-visit follow-up SMS asking for a review while the patient was still in the parking lot.
- Instagram followers: +2,100 (from 3,400 to 5,500). Not the metric that matters — but the doctor liked seeing it, so worth mentioning.
- Collected production: +$41,000/month across the practice's four ops.
- Combined MyPost Social + Glovecars fee: about 12% of that.
The measurable question for a dental practice isn't "how are my posts doing?" It's "for every DM or website inquiry, how many chairs got filled?" If the second number is under 15%, you have a response problem, not a content problem.
What we told them not to do
They wanted to run paid social. We told them wait. The organic content and the response stack had 30 days of runway to hit their capacity ceiling first — you don't spend money on paid until your organic funnel is closing at rate. By day 60 they were near hygiene capacity anyway, so paid never became necessary. If they open a fifth op we'll revisit.
They wanted us to write the doctor's Instagram captions in her voice from LinkedIn samples. We said no — the LinkedIn voice was formal, professional, DDS-signed. Instagram wants a person, not a title. We recorded four hours of the doctor talking about her practice with the office manager over coffee and modeled the captions on that. Warmer. More real. 3× the engagement.
What we did NOT do
We didn't touch Dentrix. Their existing PMS was configured fine — we just wired to it via its API. No migration risk.
We didn't take over Google Ads. They weren't running any and we told them not to start yet. Their new-patient economics were healthy on organic + response alone.
We didn't do "reputation management." The Google reviews grew because the follow-up SMS asked at the right moment. No paid review farming. No fake reviews. Just an ask, delivered when the patient was still glowing.
The point
Dental practices already have the demand. Every practice within a mile of a good ZIP has more people searching for a dentist than the practice can serve — that's not the problem. The problem is that the demand shows up as an Instagram DM at 9:47 PM on a Thursday, and by Monday morning that person has booked with the practice two blocks over.
MyPost Social pulls the demand in with actual content. Glovecars catches it before it walks away. Those are two different jobs. Most agencies do a version of the first one, badly. Almost nobody does the second one at all.
If you're paying an agency $2K-$5K/month for dental social media and your new-patient number hasn't moved in 12 months — you don't need a better agency. You need the response layer. Let me pull your last 90 days of DMs and show you the specific ones that would've booked.