Missed calls are becoming a revenue problem, not just an annoyance
Patient access begins long before a clinician enters the room. It starts the moment someone picks up the phone, and in many practices that moment is still slipping away.
Amol Nirgudkar, CPA, CGMA, estimates that 25% to 30% of calls to health care offices go unanswered during business hours. Patients who land in voicemail rarely wait around. They move on to the next practice in search results, often within minutes, and the revenue leaves with them. In a market where patients can compare options instantly through Google or chatbot tools, a missed call is not a minor service failure; it is a failed conversion point.
The clinical implication is straightforward. If access is the bottleneck, every downstream improvement in scheduling, intake, or follow-up is constrained before it begins.
Why the front desk still breaks under modern demand
The familiar front office model was built for a different era. Staffing shortages, relatively low pay for many front-desk roles, rising labor costs after the COVID-19 pandemic, and the pressure to handle more administrative work have left reception teams juggling too much at once. Some staff are trained primarily as order takers, not as triage coordinators or access specialists, and that gap shows up in the phone queue.
A practice can invest heavily in search ads and social media campaigns, then lose the patient on the first call. That is a waste of marketing spend, but it is also a systems failure. The practice paid to generate demand and then allowed the first contact to collapse. In specialties where a single visit can carry substantial per-visit revenue, the economics of an unanswered phone become hard to ignore.
The revenue range discussed in the interview was broad, running from roughly $250 to $5,000 per visit depending on specialty, which makes the opportunity cost highly variable across practices. Even at the lower end, repeated missed contacts erode margin. At the higher end, each lost appointment can represent a significant setback for a schedule that was already hard to fill.
Where AI voice agents fit without replacing human staff
Nirgudkar framed AI voice agents as an operational tool for overflow, weekend, and after-hours calls rather than a wholesale replacement for staff. That distinction is practical. Online scheduling can handle simple bookings, but telephone access still matters for patients who need clarification, reassurance, or help navigating a practice’s hours and workflow.
The best near-term use case is predictable, repetitive work. Existing patients calling to book low-acuity visits can often be routed through an AI voice agent without sacrificing service. The technology can absorb calls when the office is closed or when the queue is full, which keeps the practice from losing people who would otherwise hang up and try the next clinic. For front office teams, offloading routine calls can create more time for check-in, checkout, and the human tasks that genuinely benefit from a person at the desk.
The economics favor a narrow, targeted deployment
The pricing discussed for AI receptionist services ran from about $500 to $800 a month, with some arrangements reaching $1,000. Against that backdrop, the return depends on how many calls are saved and how much revenue each appointment can generate. Practices do not need a full-scale automation overhaul to see value. A single broken process is enough to justify a targeted intervention if that process is responsible for a steady leak in patient access.
That is why the recommendation to fix one problem at a time carries so much weight. A practice that tries to automate every workflow at once risks confusing staff, disrupting patient communication, and buying tools that never get used consistently. A narrower deployment, by contrast, lets leaders measure whether overflow calls are captured, whether after-hours access improves, and whether the front desk experiences real relief.
AI in this setting is less about novelty than throughput. A phone system that reliably answers, routes, and books can reduce friction at the exact moment patients are deciding whether to stay with a practice or abandon the search.
Patient acceptance depends on the task, not the technology label
Patients are more likely to accept AI when the request is simple and immediate. Scheduling for an established patient is a low-friction example. Insurance coverage questions are not. Those conversations often become frustrating even with a live person, because they involve plan rules, prior authorization, and benefit language that many callers already find opaque.
That means the successful deployment model is selective. AI voice agents may handle routine intake and overflow effectively, while more complex coverage discussions still need a human handoff. This is not a weakness of the technology so much as a reminder that access in medicine is layered. Patients want speed, but they also want certainty, and those two needs do not always arrive in the same channel.
A practice that understands this can reduce friction without pretending automation solves every communication problem.
Staff adoption may decide whether the purchase works
The harder challenge may not be technical at all. Getting staff to use what a practice buys often determines whether the investment pays off. A voice agent, chatbot, text tool, radiology system, or revenue cycle product can sit unused if the workflow around it never changes. Technology adoption in health care frequently fails because the people asked to use it were not involved in the design, or because the process remained broken around the new tool.
A front office team that trusts the system will route simpler calls through it and preserve its attention for patients standing at the counter. A team that views the technology as competition may work around it or ignore it. The result is the same as no purchase at all: the phone keeps ringing, patients keep leaving, and the practice keeps paying for missed opportunity.
The immediate lesson is clear. Access failures are measurable, expensive, and fixable in narrow increments. What practices need next is not a grand digital reinvention, but a disciplined approach to the single process that is bleeding patients fastest: the call that never gets answered.