Test this on a real primary-care day, not a spec sheet. A 2025 UCLA randomized trial found one ambient scribe cut average note time by about 41 seconds per note. Multiply by a 20-patient day and you get roughly 14 minutes back daily, about 4.6 hours a month. That number is the cautious one. What actually changes a day is when the notes finish, during clinic instead of after dinner. Below is the day-in-the-life math, and what it takes for the savings to actually show up.
Key takeaways
- A 2025 UCLA trial measured about 41 fewer seconds per note for one ambient tool; at 20 patients that’s roughly 14 minutes back daily, near 4.6 hours a month.
- The bigger win is timing: notes that finish during clinic instead of at 9pm. The AMA found 36 minutes of EHR time per 30-minute visit.
- The per-note saving is the floor. The visits where a scribe helps most are the long, talkative, multi-complaint ones you’d otherwise rebuild from memory.
- Against an $89 monthly subscription, recovering a fraction of an hour a day clears the cost long before you count the burnout you didn’t have.
saved per note in one tool, UCLA randomized trial (NEJM AI, 2025)
back per day at 20 patients; about 4.6 hours a month
of EHR time logged per 30-minute primary-care visit (AMA / JAMA Network Open)
What does a 20-patient primary care day actually cost you in notes?
Twenty patients is a normal primary-care day in a lot of clinics. The visits are short. The documentation isn’t. The AMA found primary-care physicians log a median of 36 minutes of EHR time per 30-minute visit, which is why the note now outlasts the appointment it describes. Stack that across twenty visits and a chunk of it lands after hours.
So the honest framing isn’t “save hours overnight.” It’s more like: shave a bit off every note, then push the rest off your evening. A 2025 UCLA randomized trial put the per-note saving at about 41 seconds for one ambient tool, from 4 minutes 30 to 3 minutes 49. Small on any single note. It adds up over a full day. And the per-note number undersells the timing shift, which is that you finish during clinic.
If you want the upstream question of how to pick a tool that delivers this, our 9-question evaluation scorecard is the place to start.
The 20-patient-day time-back model
Here’s the arithmetic, conservatively, using the verified per-note figure.
| Without an AI scribe | With an ambient AI scribe | |
|---|---|---|
| Patients per day | 20 | 20 |
| Avg note time | ~4:30 per note | ~3:49 per note |
| Documentation time per day | ~90 min | ~76 min |
| Time back per day | n/a | ~14 min |
| Time back per month (20 days) | n/a | ~4.6 hours |
| When notes finish | Often after clinic | During clinic |
Fourteen minutes a day doesn’t sound dramatic until you put it where it lands: the back of your evening. And remember the per-note figure is the cautious one. The visits where a scribe helps most are the ones you’d otherwise spend the longest reconstructing from memory afterward: the talkative patient, the multi-complaint visit, the one that loops back on itself three times before the real problem surfaces.
A primary care day: the 6pm versus the 9pm sign-off
Picture a family-medicine doctor on a 20-patient Tuesday. Without a scribe, the visits end at five and the charting doesn’t. The notes pile into the evening, and the median 36 minutes of EHR time per visit means a real share of the day’s documentation happens at home. That’s the pajama time the burnout studies keep measuring.
With an ambient scribe running, each note comes back drafted about two minutes after the visit ends. The doctor reads it, fixes what’s off, and signs, between patients or in the gaps. By six, the charts are mostly done. The 14 minutes the model shows is just the floor. What you actually feel is signing in the clinic instead of in bed.
Where the time-back model gets tested
The model assumes the scribe holds up. That’s the part to check before you trust the number. A busy clinic isn’t a quiet demo room. People interrupt. A patient buries the real complaint under three smaller ones. A relative answers half the questions. The exam room sits next to a loud waiting area. That’s where weaker tools quietly fall apart, and where the per-note saving either holds or evaporates.
So the honest test isn’t a scripted consult. It’s your busiest, loudest hour, with the patient who’s hardest to follow. If the note comes back clean from that, the 14 minutes is real. If it needs heavy cleanup, no average will rescue it.
Does the subscription pay for itself in primary care?
The arithmetic is short. Take the conservative 14 minutes a day, value it at whatever your hour is worth, and set it against $89 a month. For almost any primary-care doctor, even a fraction of that recovered time clears the subscription. And that’s before the qualitative return the burnout studies measure: fewer evenings spent charting, more eye contact while the patient is actually in front of you.
AI Medical Scribe by Patient Square is an ambient AI medical scribe that listens during the visit and hands back a structured SOAP note, ICD-10 suggestions, and a prescription draft, ready to review and sign about two minutes after the visit. The full price ladder, with no asterisks, is on our pricing page, and the way to roll it out across a small clinic is our 2-week implementation plan.
The model only gives you the floor. The real number comes from your own 20-patient day. Book a demo to see the note quality against your visit type, then run the 7-day trial on a real clinic week and count the minutes yourself.