A practice taking 60 calls a day, answering three in four during open hours and none after hours, loses about $330,000 a year in appointments that never get booked.
You take 60 calls a day, with 25% not answered live and 25% arriving after hours. About 186 calls a week reach no one, which costs you $334,081 a year in 947 appointments that never get booked.
- The same, per month: $27,840
- Calls not answered live, per year: 9,685
- Of those, after hours: 5,200
- Callers who never come back, per year: 5,443
- Appointments lost, per year: 947
- Share of all calls answered live: 53%
- Recovered by answering nearly every call, per year: $312,718
How the number is worked out.
Not every missed call is a lost patient. The model counts only the callers who never get through, and only those who wanted an appointment.
- 01Calls in. Calls per open day times open days a year. After-hours calls are added as a share of the total. At 25%, a practice with 15,600 daytime calls a year gets another 5,200 when nobody is there.
- 02Calls not answered live. Daytime calls that ring out or hit voicemail, answered callers who hang up on hold, and each after-hours call unless a person or an agent answers it.
- 03Callers who come back. Some leave a voicemail and get a callback that books. Some hang up and try again. Both are subtracted.
- 04What a lost caller was worth. Only calls that wanted an appointment count, and only the share that would have booked. A new patient is valued at Zocdoc’s first-year figure. A returning patient is valued at one primary care visit, from the 2021 federal expenditure survey.
- 05What answering nearly every call would recover. The model reruns the same practice with 3% missed, 1% lost on hold and after-hours calls answered, and takes the difference.
The defaults.
Dials you change are marked in colour. The last column says where each default comes from and what kind of source it is.
| Dial | Yours | Default | Where the default comes from |
|---|---|---|---|
| Calls per day, open hours | 60 | 60 | JRW, 2026 Our default |
| Days open per week | 5 | 5 | JRW, 2026 Our default |
| Not answered live | 25% | 25% | JRW, below CallRail and Patient Prism, 2026 Our default |
| Answered, then lost on hold | 5% | 5% | Keona Health, 2024 Vendor figure |
| After-hours share of all calls | 25% | 25% | JRW, below AnswerConnect, 2026 Our default |
| After-hours calls answered live | 0% | 0% | JRW, 2026 Our default |
| Leave a voicemail | 38% | 38% | PatientBond survey, via Solutionreach, 2025 Vendor figure |
| Voicemails that turn into a booking | 50% | 50% | JRW, 2026 Our default |
| Hang up, then get through later | 40% | 40% | JRW, 2026 Our default |
| Calls that want an appointment | 30% | 30% | OhMD, 2025 Vendor figure |
| Of those, booked when answered live | 58% | 58% | JRW case study, 2026 JRW engagement |
| New patients, of those | 35% | 35% | JRW, 2026 Our default |
| First-year value of a new patient | $538 | $538 | Zocdoc, 2023 Survey |
| Value of a returning visit | $253 | $253 | AHRQ, Medical Expenditure Panel Survey, participant publications page, 2021 Official |
Sources.
The largest measured sets are vendor platform data: CallRail’s 1.1 million leads and Patient Prism’s 11.5 million dental calls. The 42% figure quoted on many sites, from a study of 7,000 calls at 22 practices, is not a default here. None of the sites we found that quote it link to the report.
Our defaults. The booking rate is our own measurement: 58% at the practice in the voice intake case, between Liine’s 54% across 325,000 inquiries and Patient Prism’s 62% of real opportunities. Replace any default with your phone system’s report. How we set defaults.
- JRW default. JRW, 2026. No public benchmark we trust. The number is a round figure in the range we see in engagements, chosen so the headline errs low. Replace it with your own. Source
- Missed rate, set below the measured figures. JRW, below CallRail and Patient Prism, 2026. CallRail's platform data puts healthcare's missed rate at 32% of tracked leads and Patient Prism's at 31 of 100 dental callers hanging up before anyone answers. 25% is set below both so the headline errs low. Source
- Healthcare call abandonment benchmark. Keona Health, 2024. Puts healthcare call abandonment, callers who hang up on hold, at 5% to 7%. Source
- After-hours share, set below the measured figure. JRW, below AnswerConnect, 2026. AnswerConnect's call logs put 47% of calls to US businesses outside Monday to Friday working hours, across every industry and among businesses that already pay for after-hours answering. 25% is set well below it for a practice with fixed hours. Source
- Patients who reach voicemail and hang up. PatientBond survey, via Solutionreach, 2025. 62% of patients who reach voicemail hang up without leaving a message. Repeated across vendor pages; the survey itself was not located. Source
- Why patients call: reasons by share of calls. OhMD, 2025. Scheduling and rescheduling 25% to 30% of calls, refills 15% to 20%, confirmations 10% to 15%, billing 5% to 10%. The vendor's own analysis across its practices, method not published. Source
- Voice agent at a mental health practice. JRW case study, 2026. Booking rate for answered new-patient calls rose from 33% to 58% over six weeks at a practice with over 100 providers. One engagement. Liine's 54% across 325,000 inquiries and Patient Prism's 62% of real opportunities sit either side of it. Source
- New patient value by practice size and specialty. Zocdoc, 2023. 351 medical groups and 11 health systems surveyed by a scheduling marketplace. First-year value of a new patient: $538 at medical groups, $1,395 at health systems. Source
- Mean expense per office-based physician visit, by specialty. AHRQ, Medical Expenditure Panel Survey, participant publications page, 2021. 2021 mean expense per visit: $253 for primary care, $368 across all specialties, $433 OB/GYN, $436 cardiology, $627 orthopedics. The federal survey of what visits cost, all payers combined. Source
Also consulted
- Dental Patient Access Report. Patient Prism, 2025. 11,552,668 calls to 8,280 dental locations on the vendor's platform in 2025. Per 100 calls: 31 hung up before reaching anyone, 34 were real booking opportunities, 21 booked, 13 walked away, 4.8 got a follow-up call and 0.7 of those came back. Dental, not medical, and one vendor's customers. Source
- Missed-call rate by industry, 1.1 million tracked leads. CallRail, via Plumber Magazine, 2025. Call-tracking platform data, January 2025: healthcare 32% of leads missed, legal 28%, home services 14%, real estate 9%. Lead-level, marketing-generated calls only, so it understates total call volume and overstates how many were new business. Source
- 325,000 new-patient calls and forms, booking rate. Liine, 2022. 54.1% of new-patient inquiries booked an appointment; a 2023 cut of 278,000 marketing-sourced leads booked at 39.9%. Call-tracking platform data. Source
- 47% of business calls occur after hours. AnswerConnect, 2026. US call logs, 5,906,476 minutes, April to June 2026: 47% of calls arrived outside Monday to Friday working hours or on weekends, across all industries. An answering service's own customers, who skew toward businesses that already get after-hours calls, so the defaults here sit well below 47%. Source
- 7,000 calls across 22 medical practices in 18 states. Circulated by notifyMD, TrackableMed and AnswerNet, 2025. 42% of incoming calls missed. Widely cited; we could not locate the original report or its definition of 'missed'. Treated as the upper bound. Source
- Legal Trends Report, secret-shopper study. Clio, via the American Bar Association, 2024. 500 US firms phoned as prospective clients: about 40% answered live, down from 56% in 2019; about 52% made contact once callbacks are counted. Source
Where the calls go.
-
The front desk is on another call
Most daytime misses happen while staff are on the phone or with a patient. A voice agent picks up the second line, reads the schedule in your EHR and books while the caller is on the line. How our voice agent answers for practices.
-
After hours
A quarter or more of calls arrive after hours. The default assumes none are answered live. An agent that books on evenings and weekends closes the largest leak in the model.
-
On hold too long
Hold loses fewer callers than the other two leaks. An agent that takes refill, directions and confirmation calls off the desk shortens the queue for everyone else.
One mental health practice went from answering one new-patient call in four to all of them. The case.
Questions.
- Is 25% missed realistic?
- It is below the measured figures. CallRail’s 2025 analysis of 1.1 million leads put healthcare at 32%. Patient Prism found 31 of every 100 callers to 8,280 dental offices hung up before reaching anyone. One multi-practice analysis claims 42%. Your phone system report has your own rate. Put it in.
- Why isn’t every missed call lost?
- Many callers leave a message and get called back. Many hang up and try again an hour later. The model subtracts both first.
- Where do the dollar values come from?
- A new patient is $538, Zocdoc’s first-year figure across 351 medical groups. A returning patient is $253, the federal survey’s 2021 mean cost of a primary care visit. Specialties run higher: $433 OB/GYN, $436 cardiology, $627 orthopedics. Health systems should use Zocdoc’s $1,395.
- We have an answering service. Does that count?
- Only if it books the appointment. A service that only takes a message leaves the caller waiting for a callback, as voicemail does. Set “after-hours calls answered live” to the share it schedules.
“Copy a link” puts your dials in the address, and we store nothing. A browser agent that supports WebMCP can run this calculator as estimate_missed_calls_medical_practice_cost. How the calculators work.
Answer the calls your front desk can’t.
Bring your phone system report and tell us your EHR, and we’ll show you which calls a voice agent can book. Fixed price per automation. If the first one misses the scope we agreed, you don’t pay.
Book a call (opens in a new tab)Related: Missed call statistics · How to measure missed calls · AI voice agents and HIPAA · AI automation for medical practices · Voice intake case study