Pain-Free Dental Marketing helps dental practices across the country grow their new-patient base, and every campaign it runs is judged on one number: how many of the calls it generates end up on the schedule.
For years, the agency got that answer the hard way, by manually listening to calls. An internal team, nicknamed the "CallRail Army," listened to every call and tagged it by hand as a new patient, existing patient, or spam. At its peak, roughly eight full-time employees were reviewing about 20,000 calls a month across the agency's whole client base.
The process delivered the data clients needed, but it required significant time and resources to maintain. Pain Free Dental Marketing needed a more scalable way to get the same level of insight without manually reviewing every call.
Here's how the agency used CallRail's Call Tracking to automate call tagging with AI and what changed for its dental clients once they could get accurate conversion data faster.
Using CallRail, Pain-Free Dental Marketing now:
- Identifies new patients with 94% accuracy and existing patients with 98% accuracy
- Catches inflated schedule-rate numbers before they reach a client, isolating the real rate at 58.3%
- Separates real new-patient calls from wrong numbers, spam, and existing patients calling in on branded search terms
- Turns tagging around same-day across 20,000 monthly calls, down from a multi-week lag
- Sees which channels convert new patients and which don't, a gap that used to be invisible
When call volume outgrew a manual listening team
The CallRail Army started as a side project roughly four to five years ago, and it started because nobody else was going to do it. "We were kind of doing it on our dime," says Eric Hubbard, Pain-Free Dental Marketing's co-founder.
As call volume grew, tagging lagged by a couple of weeks, so client teams worked off stale data. Some calls were wrong numbers — callers who'd clicked one dentist's ad and reached another practice entirely. Existing patients calling in on branded search terms got counted as new leads, inflating the number that mattered most to clients. And a call that rolled to voicemail often didn't register as missed, even though an unreturned voicemail is a lost new-patient opportunity.
The agency tried automating this once before, well before this rebuild. It missed too many new-patient calls to be trusted, and the team went back to doing it by hand.
A missed tag costs more than a missed call
A new-patient case can run $2,000 to $30,000 or more, Eric says. And the numbers add up fast. Closing at 23% instead of a healthy 60% roughly doubles what every new patient costs to acquire, on exactly the same ad spend. Get that tag wrong, and a practice loses more than a lead. It costs a patient relationship, or triggers a conversation about ad spend that has nothing to do with what's actually happening at the front desk.
And it's not a rare problem. Businesses miss an average of 28% of their calls industry-wide.
The manual team's cost and headcount weren't sustainable at the agency's scale, and the real constraint was never the ad budget. It was whether anyone answered the phone and got someone scheduled.
More ad spend won’t help if the phone isn’t being answered, or patients aren’t being scheduled in for treatment.
– Eric Hubbard, Co-Founder, Pain-Free Dental Marketing
That meant the agency needed to know what was actually happening once someone picked up the phone.
How the system works now
The agency built a custom stack on top of CallRail. Call Tracking captures and transcribes every inbound call.
From there, a layer is added through Weaviate, which pulls the transcript data into BigQuery. AI models read each transcript and tag the call as new patient, existing patient, or scheduled — Claude Sonnet handles new-patient detection, while Gemini Flash and Pro handles existing-patient detection.
Verified tags get pushed straight back into CallRail. A much smaller QA team now reviews only the calls the models flag as uncertain, instead of listening to all 20,000. What used to take a multi-week lag now happens same-day.
The agency never built a separate dashboard for the new system. Clients and coaches see the same tags, on the same call, inside the same CallRail record the agency is already looking at. Nobody has to learn a new tool, and nobody has to take agency-only numbers on faith.
What one dental practice's numbers looked like once the data could be trusted
By May 2026, the CallRail-fed pipeline was producing real numbers to show for it.
-
1,030 total calls, 242 new patients, and 141 scheduled appointments for the month
A new-patient schedule rate of 58.3%, corrected after the agency caught earlier reporting that had inflated it
The same schedule-rate view across every channel the agency runs, from 67.3% for Google Business Profile down to 60.7% for the website pool, and 43.2% for Google Ads
That 58.3% lands close to the 60%-plus booking rate that marks a healthy practice. For the first time, this client had a specific number to aim for instead of a vague sense of "doing better."
These numbers provided better clarity on how to effectively make marketing changes. For example, the team was able to adjust paid advertising efforts by including negative keywords to cut Medicaid-adult spend that wasn't converting, a new campaign for emergency and same-day searchers the existing setup wasn't able to catch, and an audit into why Google Ads kept underperforming.
But the ad account was the smaller half of the story. In the same month, 101 new patients called and never got scheduled. A sample of 100 transcripts showed why, and none of the reasons were marketing: confusion about Medicaid eligibility, uncertainty about whether the practice took their insurance, and an AI phone agent that couldn't transfer callers to a human when they asked for one.
Beyond the paid advertising fixes, the team also brought more clarity to these questions directly on the practice's website, helping filter out callers who weren't the right fit before they ever picked up the phone.
A clearer read on new patients
Pain-Free Dental Marketing talks to clients differently now. Before, client conversations would open with the agency asking for more budget. Now it starts with what's actually happening at the front desk. The agency can see the difference between a real new-patient call, a wrong number, a missed voicemail, and an existing patient calling in on the wrong ad.
Both sides can pull up the same call and talk through it together, because everything lives in the same CallRail record instead of a private dashboard only the agency can see. As Eric puts it, "it's just data.
Other dental coaching partners have already asked about rolling the same system out across their own accounts. The agency is exploring turning it into a standalone product beyond its own marketing clients.
Turn your own calls into data you can trust
Marketing can only ever amplify what already happens when someone picks up the phone. Now we can see both. CallRail helps you see exactly which calls are new patients and which get scheduled, on the same record your clients already see.
