From 62% to 94% inquiry capture in 45 days, without replacing a single staff member

The practice

Maria runs a four-clinician group practice in suburban Atlanta. She has been in the field for fourteen years, two of them as a practice owner. By most measures, the practice was doing well: a consistent referral stream, strong clinical outcomes, and a front desk coordinator who had been with her for three years.

She did not think she had an intake problem. The practice was busy. There was usually a short waitlist. If inquiries were slipping through, it was not obvious.

She was right that the practice was busy. She was wrong about why.

The measurement

See what this looks like for your practice

In 30 minutes we will identify where your practice is losing clients, quantify the revenue impact, and give you the single highest-leverage fix to implement first.

Book a Strategy Call →

During the initial assessment conversation, we asked Maria three questions: What is your inquiry capture rate? What percentage of calls come in after 5pm? What is your average response time for after-hours inquiries?

She did not know the answers to any of them. That is normal. Most practice owners do not have those numbers because their systems do not surface them automatically.

We pulled 60 days of data from her phone system logs, email inbox, and EHR. The numbers:

Total inbound inquiry attempts: 87 across 60 days
Captured as a contact record: 54 (62% capture rate)
Inquiries arriving after 5pm or on weekends: 31% of total
Average response time for after-hours inquiries: 22 hours

Of the 33 inquiries that were not captured, 28 had come in after business hours. They went to voicemail. Some left messages; most did not. None of them were called back within two hours. Most were not called back the same day.

Maria's practice was not at capacity. It was losing a third of its potential demand before any conversation started.

What the 33 uncaptured inquiries meant

At a 62% capture rate and a 52% conversion rate on captured inquiries, Maria's practice was converting approximately 11% of total inbound demand to first appointments. The math for a four-clinician practice at those rates looked like this: 87 inquiries per 60 days, 11% end-to-end conversion, 9 to 10 new first appointments per 60 days.

Raising capture rate to 90% and holding conversion constant would produce 16 to 17 first appointments in the same window -- a 70% improvement in new client throughput with no change in marketing spend, no new clinicians, and no change to the existing intake process for business-hours inquiries.

That is what was sitting in the after-hours voicemail gap.

The change

Implementation focused on one thing: ensuring that every inbound inquiry got a response within three minutes, regardless of when it arrived.

An AI voice layer was configured to answer all inbound calls outside business hours and overflow calls during peak periods. The configuration for Maria's practice prioritized warmth over efficiency: the voice assistant introduced itself by name, asked what had brought the caller in, and offered either immediate scheduling or a morning callback from the coordinator, depending on what the caller preferred.

For email and web form inquiries arriving after hours, an automated response went out within five minutes with specific next steps and a direct scheduling link.

Nothing changed about how the coordinator handled business-hours inquiries. The coordinator's role expanded to include reviewing the overnight contact log each morning and following up on anything that had not self-resolved through the automated scheduling flow.

45 days later

At the 45-day mark:

Inquiry capture rate: 94% (up from 62%)
Conversion rate: 59% (up from 52%, modest improvement)
Average response time, after-hours inquiries: 4 minutes (down from 22 hours)
New first appointments: 16 in 45 days (compared to 9 to 10 in the previous 60-day window)

The conversion rate improvement was smaller than the capture rate improvement, which is typical. Capture is faster to move because it is a systems gap, not a relationship gap. Conversion depends on factors that take longer to optimize: the quality of the first response, the scheduling experience, and how well the initial intake conversation sets expectations.

What moved immediately was volume. Maria's practice went from losing a third of its demand before any contact was made to recovering almost all of it.

What Maria said

"I thought we were missing calls because people were not leaving voicemails. It turns out we were missing people who were not going to leave a voicemail -- because why would they, at 8pm on a Wednesday, when they could just try somewhere else? The data was embarrassing to look at. But it was also a relief, because I finally understood what was happening."

Maria, Practice Director

Frequently asked questions

How do you define "captured" versus "not captured" for an inquiry? +

A captured inquiry is one where the practice has a logged contact record with enough information to initiate follow-up: at minimum a name and a contact method (phone or email). An inquiry is not captured if the person called, did not leave a voicemail, and no outbound contact was initiated. It is also not captured if a voicemail was left but no outbound contact was made within a reasonable window. In Maria's case, we pulled phone system logs showing total inbound call attempts and compared them against contact records in the EHR. The gap between those two numbers is the uncaptured inquiry count.

What happens to after-hours callers who prefer to speak with a human? +

They are offered a callback. The voice assistant explicitly gives callers the choice: schedule now using the automated flow, or be added to the morning callback queue for a call from the coordinator during business hours. The majority of callers who prefer human contact choose the callback option rather than abandoning the inquiry entirely -- which is what they were doing before when they got a generic voicemail.

What if the practice uses a specific EHR that may not integrate with the AI system? +

BHSAI systems are configured to work with the intake and scheduling workflow that already exists in the practice. For most practices, the AI layer handles the pre-EHR portion of intake -- inquiry capture, initial information collection, and scheduling -- and creates a handoff record that feeds into the EHR at the point where the EHR would normally be used. Full integration with SimplePractice, TherapyNotes, Jane App, and Athenahealth is supported. Other systems are assessed during discovery.

What is the minimum practice size for this kind of implementation to make financial sense? +

The ROI calculation is based on inquiry volume and session fee, not practice size directly. A solo practitioner with a high inquiry volume and a strong referral network may have a stronger case than a two-clinician practice that is at capacity and has a long waitlist. The break-even point is roughly: if the system captures 3 additional clients per year who would otherwise have been lost, the investment pays for itself at typical behavioral health session fees. For most practices receiving more than 15 to 20 new inquiries per month, that threshold is crossed well before year one.

How long before the system runs without needing active management? +

The AI layer runs without active management from day one -- that is the point. The coordinator's role shifts to reviewing the overnight log and handling exceptions, not monitoring the system actively. The first two weeks involve daily check-ins to refine the configuration based on real call patterns. After 30 days, most practices move to a weekly review. After 90 days, the system runs effectively as a background process that the coordinator interacts with for exceptions only.

What did the coordinator think about the change? +

Maria's coordinator was initially skeptical, which is common. The concern is usually about whether the AI will say something wrong to a client or handle a sensitive situation poorly. After two weeks of reviewing the overnight logs and seeing how the system handled calls, her coordinator's response was: "I wish we had done this two years ago." The work that had previously consumed her mornings -- returning voicemails, chasing after-hours callbacks -- was handled before she arrived. She spent the time on client communication and scheduling complexity instead.

Did the quality of new clients change with higher capture rates? +

Not in any way that was clinically significant. The concern some practice owners raise is that clients who reach out after hours or who are captured by an automated system are somehow less serious about treatment. The data from Maria's practice does not support that. Retention rates for clients who came through the AI-assisted intake were comparable to those who came through the human-only process. The clients who were previously lost were simply people who called at 8pm instead of 10am -- not a different population, just a different time.