The missed-call cost model: from phone log to dollars
Your phone system already knows how many calls went unanswered. What it cannot tell you is what those calls were worth. This model walks a month of missed calls through a simple funnel, so you end up with a dollar figure you can defend and a per-call value you can track.
- Start from unique, unresolved missed calls, not the raw missed-call count.
- Split callers by intent: new patients, existing patients trying to book, and everything else.
- Value new patients by their first-year visits, and existing patients by the visit at stake, and keep the two on separate lines.
- Dividing the total by raw missed calls gives a value per missed call, a simple number to track month over month.
- When calls are missed matters: lunch, early morning and after-hours calls often need a different fix from mid-morning overflow.
Step 1: from raw missed calls to unresolved callers
Export one month of call activity from your phone system. Most systems report total inbound calls, answered calls, and calls that rang out, went to voicemail or were abandoned. Then narrow the list down, because the raw count overstates the problem.
Here is an illustrative practice. None of these numbers are benchmarks; they show the method.
- Inbound calls: 2,400 in the month. Answered: 82%, so 432 missed.
- Remove repeats and very short hang-ups. Several missed calls often come from the same number, and calls abandoned in the first few seconds are frequently misdials. Say that leaves 300 unique missed callers.
- Remove callers who were later reached. Match the missed numbers against answered calls and outbound callbacks within 24 hours. Say 120 were reached, leaving 180 unresolved callers.
Step 2: split by caller intent
Not every unresolved caller wanted a visit. Estimate the mix by listening to voicemails for a week, or by having the desk tag the reason on every answered call for two weeks and assuming missed calls look similar. In this illustrative practice:
- New patients: 20%, or 36 callers.
- Existing patients trying to book or rebook: 45%, or 81 callers.
- Everything else (billing, records, refills, vendors, other offices): 35%, or 63 callers. These matter for service, but they are not counted as lost visit revenue in this model.
Step 3: put a value on each group
New patients. A new patient is worth more than one visit, because a patient who books usually returns. Use first-year value rather than lifetime value, which is too speculative. Illustrative example: a first visit worth $210 plus two follow-ups at $140 gives a first-year value of $490.
Not every new caller would have booked even if answered. Assume an illustrative 60% would have. 36 x 60% = 21.6 patients. 21.6 x $490 = $10,584 in first-year value.
Existing patients. Many existing patients call back later or book through another route, so only part of this group is lost. Assume an illustrative 30% end up not booking, or booking much later. 81 x 30% = 24.3 visits. 24.3 x $140 = $3,402.
Step 4: total it and find your per-call value
Add the two lines: $10,584 + $3,402 = $13,986 for the month in this illustrative example. Report the two lines separately, because the new-patient line includes follow-up visits that happen later in the year.
Now divide by the raw missed-call count from Step 1: $13,986 / 432 = about $32 per missed call. This one number is useful because it moves with everything that matters. If the answer rate improves, missed calls fall. If callbacks get faster, unresolved callers fall. If the new-patient share rises, the value per call rises. Track it monthly alongside the answer rate.
Notice what the model does not include: the cost of the staff time to return calls, and the reputation cost of callers who give up. Both are real but hard to measure, so leaving them out keeps the estimate conservative.
Step 5: look at when calls are missed
Break missed calls down by hour and day. The pattern points to the fix.
- Opening rush (first hour). Many patients call as soon as the office opens. A spike here usually means too few people on phones at the busiest moment.
- Lunch. If the phones go to voicemail while the desk eats, every lunch-hour caller is in the missed column. Staggered lunches or an automated text-back help.
- Mid-day overflow. When check-in and phones share the same people, phones lose. The patient standing at the desk always wins.
- After hours and weekends. Callers here expect less, but they still choose whoever responds first in the morning. A prompt text with open times can reach them before they call another office.
What speed does to the model
The model assumes a caller is either reached or not. In reality, the value of a callback falls the longer it takes, because the caller may already have booked elsewhere. A useful habit is to record the time from missed call to first response and report the median each week.
A text-back script a front desk could use, kept to minimum necessary information: "[Practice name]: sorry we missed your call. Reply 1 to book a visit, 2 for an existing appointment, 3 for billing or records. Reply STOP to opt out." It works after hours, does not depend on the caller leaving a voicemail, and turns a missed call into a sorted request.
DocsMD turns missed calls into booked visits, and proves it
DocsMD detects missed calls and responds around the clock with a minimum-necessary text: practice name, open times and numbered choices, never the patient's name or visit reason, with STOP honored on every message. Bookings are handed to your front desk to enter, and every recovered visit appears in your weekly report with the patient, date, action and dollar amount.
Questions
Why not just multiply missed calls by my average visit value?
Because many missed calls are repeats, misdials, callers you reached later, or calls that were never about a visit. Multiplying the raw count by visit value can overstate the loss several times over.
Should I use lifetime value for new patients?
First-year value is safer. Lifetime estimates depend on assumptions that are hard to verify, and a conservative figure is easier to defend when you make staffing or spending decisions.
How do I estimate caller intent if nobody leaves voicemails?
Have the front desk tag the reason for every answered call for two weeks. Missed calls tend to follow a similar mix, though new patients may be slightly more likely to hang up without leaving a message.
Should calls from existing patients count as lost revenue at all?
Only partly. Most existing patients try again or reach you another way, which is why this model counts only an assumed share of them as lost. Replace that assumption with a measured figure by checking how many unresolved existing-patient callers had no visit booked in the following two weeks.
What is a good answer rate?
Rather than chase an outside number, track your own answer rate and value per missed call each month. Both should move in the right direction as you fix staffing, lunch coverage and after-hours response.