Patient no-shows have been creeping up, not down. MGMA’s DataDive benchmarking put the median medical group no-show rate at roughly 6.81% in 2023, up from about 5% the year before, and 37% of groups reported no-shows rising again the following year. For a 3-provider clinic seeing 25 patients a day per provider, that swing is the difference between two or three empty chairs a week and nearly one a day. The number that matters isn’t a generic per-appointment estimate — it’s your own visit value multiplied by your own no-show rate. The fix is a reminder-and-waitlist loop, not a bigger front desk.
Why no-shows are worse than they look
A no-show doesn’t just cost the missed visit. It costs a slot that could have gone to a patient waiting for an opening, plus the provider and room time that already got allocated and can’t be resold once the hour passes. Unlike a cancellation with notice, there’s usually no chance to backfill the slot at all.
That’s why the rising trend matters more than the absolute number. MGMA’s 37% figure — more than a third of medical groups seeing no-shows climb further in 2024 — means this is a growing leak for most practices, not a one-time blip tied to a bad flu season or a stretch of bad weather.
Why this leak hides from the schedule
The schedule always looks full on paper, so the leak only shows up as revenue that never arrives. A clinic’s calendar can show every slot booked and still bleed money, because a filled slot and a kept appointment are two different things. Owners tend to notice no-shows only when they’re glaring — a provider sitting idle for an hour — and miss the steady 1-in-15 pattern that’s costing far more over a quarter.
The other reason it hides: no-show data usually lives in the scheduling system, not the P&L. Nobody runs a monthly report that says “here’s what empty chairs cost us this quarter,” so the number never gets named, and what doesn’t get named doesn’t get fixed.
Build the math from your own visit value
Skip any borrowed “cost per missed appointment” figure — practices vary too much in payer mix and visit type for a generic number to mean anything. Build it from four inputs you already know:
| Input | Example |
|---|---|
| Providers | 3 |
| Visits scheduled per provider, per day | 25 |
| No-show rate | 6.81% (MGMA 2023 median) |
| Average visit value (net collected, not billed) | $140 |
Three providers × 25 visits × 6.81% works out to roughly 5 no-shows a day across the clinic. At $140 average visit value, that’s about $700 a day, or roughly $16,000 a month in visits that were scheduled, staffed for, and never delivered — before counting the wasted room and provider time.
Run your own numbers with your real visit value and your real no-show rate (pull it from your scheduling system rather than guessing — most owners underestimate it). If your rate is closer to 3%, this still adds up over a year; if it’s north of 10%, this is very likely your biggest leak.
The fix: reminders and a live waitlist, not more front desk
Most of this gap closes with plain automation, not judgment calls:
- Multi-touch reminders. A single reminder the day before catches some patients. A sequence — confirmation at booking, reminder at 48 hours, reminder at 24 hours, each with an easy reschedule link — catches more, because it gives patients more chances to flag a conflict early enough for you to fill the slot.
- A live waitlist, not a paper one. When a patient cancels or reschedules, the freed slot should offer itself to the next waitlisted patient automatically, rather than waiting for someone at the front desk to remember to call down a list.
- A no-show pattern flag. Patients who no-show repeatedly get a different, more direct outreach — a phone call instead of a text — because the automated sequence isn’t reaching them.
Where AI earns a place here is narrow: drafting the reminder and waitlist-offer messages in a patient’s preferred channel and tone, and flagging the repeat no-show pattern for staff attention. It never decides who gets bumped, never cancels or rebooks a visit on its own, and everything in this loop runs inside HIPAA-compliant tools under a signed BAA. A human approves anything that actually touches a patient’s schedule or record.
When the honest answer is more front-desk staff
If your no-show rate is high because patients genuinely can’t get through to reschedule — phones ringing out, hold times long enough that people give up — that’s a staffing gap, not a workflow gap, and automation won’t fix it. The same is true if your patient population skews toward transportation or insurance instability that no reminder text solves; that calls for care coordination or a social-work referral relationship, not another SMS.
And if you’re a single-provider practice with a tight, known patient panel, you may already be catching most of this by instinct — measure first before assuming there’s a gap to close.
Find your clinic’s number
This is the cash leak — the one that shows up as a schedule that looks full but a bank balance that doesn’t match. The 3-minute scorecard also scores the other two: the clinical hours going to prior authorization instead of patients, and the new patient who called twice and booked somewhere else. Free, no call, no pitch — see how these leaks show up across use cases or run your own number now.