Four ways AI can help practices reach the patients they’re missing and get paid for care already delivered. Together, they’re worth up to $125,000 (or 10%) per physician a year, without adding staff.
One in six calls going unanswered during office hours is lost revenue. Front-end denials from inactive plans and misspelled names are the cheaper part of that gap to close first.
The 18% daytime missed-call figure is the more interesting number than the revenue estimate, because it's an operations problem a practice can measure this quarter. The question I'd want answered is whether the recovered bookings come from genuinely new patients or from rescheduling people who would have called back anyway, since only one of those is incremental.
Agree, the missed-call number is a useful one because any practice measures its own missed-call rate this way. On incrementality: the model doesn’t count every missed call as lost. Our data shows that about 20% of callers who wanted to book will call back and end up booking.
Healthcare has discovered the abandoned cart, except the cart contains a patient, an insurer and three definitions of "covered." AI can recover a lot of operational leakage; the harder question is which friction was failure and which friction was information.
I see Lever 3 (Protecting Reimbursements) as targeting the friction that is a clear solvable failure: when denials occur because of billing a plan that ended last month or a missing approval nobody checked for, or, as happened to me, a misspelling of a name. When an insurer says no for a real reason, the goal is to find out WHY before the visit, not after, so that care doctors provide is paid for ultimately.
A booked appointment is a promise, not a receipt. That $1.3 million counts slots filled, not visits attended or claims paid — and the post's own figures say 18.8% of appointments get missed on average and 10% of first claims come back denied. Worth reading twice, too, that the research and the calling software come from the same company.
The reported 18% daytime missed-call rate is more actionable than the headline estimate. Did CallIQ show whether that problem varies materially across specialties?
Or…AI just eliminates physicians as we know it today, develops new genetic and drug treatments to eradicate ills and thus reduces Medicare and insurance costs.
3 is the only solid opportunity here. 1) there is a pretty large assumption here that the 1 in 6 who hang up before the call is answered don't call back under regular circumstances. Also that an AI answering their call won't have a negative impact on likelihood to recommend practice to other patients. 2) there is an assumption that the patients who don't return need or want to return just because the chart says they are "due". What does that actually mean? 4) that AI collection calls aren't going to negatively impact likelihood to recommend practice to other patients.
And the biggest assumption - that applying the same fix will work across different specialities. Looking forward to seeing what Prosper AI can accomplish. Regardless of the assumptions I think it's good that someone is doing something to address the access problem.
1. The model assumes about 20% of missed callers who wanted to book call back anyway, and only counts the rest. On whether patients mind an AI answering: It’s interesting! We have so much data about people (especially older people) really enjoying talking to AI, and even sharing positive reviews mentioning AI agents by name on Google, etc.
2. “Due” means the patient was told to come back, for example for an annual skin check or for a follow-up after surgery. Unless that next visit is booked before they leave, some of those visits never happen.
3. The calls aren’t collections in the usual sense. They explain the balance, answer questions about charges and insurance, and offer a payment plan or a person in billing. Patients who understand what they owe are more likely to pay it.
On specialties, agreed that there will be differences in how specialties apply the fixes and that some levers may be more important than others depending on the specialty. That’s why we created a calculator where specialities can quantify the impact they can see. https://www.getprosper.ai/practice-revenue-growth-calculator
Great to see you're adjusting for bad call abandonment versus callbacks. 20% is reasonable. The next big challenge for improving access... Referral distribution!
One in six calls going unanswered during office hours is lost revenue. Front-end denials from inactive plans and misspelled names are the cheaper part of that gap to close first.
The 18% daytime missed-call figure is the more interesting number than the revenue estimate, because it's an operations problem a practice can measure this quarter. The question I'd want answered is whether the recovered bookings come from genuinely new patients or from rescheduling people who would have called back anyway, since only one of those is incremental.
Agree, the missed-call number is a useful one because any practice measures its own missed-call rate this way. On incrementality: the model doesn’t count every missed call as lost. Our data shows that about 20% of callers who wanted to book will call back and end up booking.
Healthcare has discovered the abandoned cart, except the cart contains a patient, an insurer and three definitions of "covered." AI can recover a lot of operational leakage; the harder question is which friction was failure and which friction was information.
I see Lever 3 (Protecting Reimbursements) as targeting the friction that is a clear solvable failure: when denials occur because of billing a plan that ended last month or a missing approval nobody checked for, or, as happened to me, a misspelling of a name. When an insurer says no for a real reason, the goal is to find out WHY before the visit, not after, so that care doctors provide is paid for ultimately.
Fair. A misspelled name becoming a healthcare financing event is exactly the kind of friction AI should kill.
A booked appointment is a promise, not a receipt. That $1.3 million counts slots filled, not visits attended or claims paid — and the post's own figures say 18.8% of appointments get missed on average and 10% of first claims come back denied. Worth reading twice, too, that the research and the calling software come from the same company.
The reported 18% daytime missed-call rate is more actionable than the headline estimate. Did CallIQ show whether that problem varies materially across specialties?
Or…AI just eliminates physicians as we know it today, develops new genetic and drug treatments to eradicate ills and thus reduces Medicare and insurance costs.
As a child of two doctors, with tremendous respect for the medical profession, this is not a future I want to see.
3 is the only solid opportunity here. 1) there is a pretty large assumption here that the 1 in 6 who hang up before the call is answered don't call back under regular circumstances. Also that an AI answering their call won't have a negative impact on likelihood to recommend practice to other patients. 2) there is an assumption that the patients who don't return need or want to return just because the chart says they are "due". What does that actually mean? 4) that AI collection calls aren't going to negatively impact likelihood to recommend practice to other patients.
And the biggest assumption - that applying the same fix will work across different specialities. Looking forward to seeing what Prosper AI can accomplish. Regardless of the assumptions I think it's good that someone is doing something to address the access problem.
Fair challenges, taking them in order.
1. The model assumes about 20% of missed callers who wanted to book call back anyway, and only counts the rest. On whether patients mind an AI answering: It’s interesting! We have so much data about people (especially older people) really enjoying talking to AI, and even sharing positive reviews mentioning AI agents by name on Google, etc.
2. “Due” means the patient was told to come back, for example for an annual skin check or for a follow-up after surgery. Unless that next visit is booked before they leave, some of those visits never happen.
3. The calls aren’t collections in the usual sense. They explain the balance, answer questions about charges and insurance, and offer a payment plan or a person in billing. Patients who understand what they owe are more likely to pay it.
On specialties, agreed that there will be differences in how specialties apply the fixes and that some levers may be more important than others depending on the specialty. That’s why we created a calculator where specialities can quantify the impact they can see. https://www.getprosper.ai/practice-revenue-growth-calculator
Great to see you're adjusting for bad call abandonment versus callbacks. 20% is reasonable. The next big challenge for improving access... Referral distribution!