There's a number I keep coming back to in conversations with practice leaders: almost 40%.
That's the share of inbound referrals across specialty healthcare that never convert to a booked appointment. Not because patients don't need care. Not because referring providers send bad referrals. Because the operational infrastructure between receiving the fax and contacting the patient can't keep pace with the volume.
Thousands of patients a month, already referred, already in the system, just disappearing from the pipeline. And with them, millions in annual revenue that the practice had already earned the right to capture.
I've spent 25 years selling into healthcare, from the Palm Pilot era at Epocrates through WebMD's mobile launch and Jumo Health's PE exit. And I can tell you: this is one of the most expensive operational blind spots I've encountered. Not because it's hidden (most practice leaders know their referral conversion isn't great) but because very few have actually done the math on what that gap costs them.

The Follow-Through Problem

Let's be specific about where the breakdown happens.
A referring physician sends a fax. That fax lands in a queue, sometimes digital, often still paper. Someone at the front desk needs to read it, extract the patient demographics, verify insurance, create a record in the EHR, and then call the patient to schedule. In a practice receiving 7,000 referrals a month, that's an enormous manual workload layered on top of phones that are already ringing off the hook.
The result is predictable: prioritization by triage. The referrals that get worked are the ones that happen to be on top of the pile, or the ones where the referring physician's office calls to follow up. The rest wait. And wait. And eventually go stale.
This isn't a staffing failure. It's a structural one. You could hire more people, and you'd still have the same bottleneck, because the constraint isn't effort. It's the sequential, manual nature of the workflow itself.

Quantifying the Gap

Here's where the conversation gets uncomfortable for most practice operators.
At a 40% referral loss rate, a practice receiving 7,000 referrals a month is losing roughly 2,800 patients per month who were already referred and never scheduled. At an average of $150 per visit, that's approximately $420K in monthly revenue leakage, or nearly $5M annualized.
That's not a projection. That's arithmetic based on real referral volume patterns we see across large specialty practices.
Now compare that to a benchmark: practices running fully autonomous AI referral processing convert at nearly 90%. Same patient populations. Same referral sources. Same payer mix. The difference is entirely in the operational infrastructure between fax receipt and booked appointment.
Closing that gap, from 60% conversion to nearly 90%, represents millions in recoverable annual revenue. Not from new patient acquisition. Not from expanded marketing. From patients who were already sent to you.

Why This Blind Spot Persists

In my experience, there are three reasons practice leaders underestimate this problem.
First, referral conversion isn't a metric most practices track with precision. They know how many referrals come in. They know how many appointments they book. But the connection between those two numbers, the conversion funnel, often lives in a spreadsheet that gets updated quarterly at best.
Second, the cost is invisible on the P&L. Revenue leakage from unbooked referrals doesn't show up as a line item. It shows up as the gap between where revenue is and where it could be. And that gap is easy to attribute to "market conditions" or "seasonality" rather than operational throughput.
Third, the traditional solution (hire more staff) has diminishing returns. You can add bodies to the phone bank, but you can't fundamentally change the throughput of a sequential, manual fax-to-booking process by adding more people to it. The bottleneck is architectural, not staffing.

What Changes the Math

What we deployed was end-to-end autonomous referral processing. AI receives the fax, extracts demographics, diagnosis, and referring provider. Creates the patient record directly in the EHR. Contacts the patient via phone and SMS. Matches availability, insurance, and location preference. Books the appointment and confirms it in the scheduling system.
No handoffs. No staff involvement at any step. One hundred percent autonomous.
The results were immediate and measurable: conversion jumped to nearly 90%, with thousands of appointments booked in the first weeks of deployment, all without a single staff member touching the workflow.
But the number that matters most to the PE operating partners and COOs I talk to every day isn't the conversion rate. It's the annualized revenue impact. For a practice doing 7,000 referrals a month, the difference between a 60% conversion rate and a 90% conversion rate is roughly $3M+ in annual revenue. And for practices starting closer to 50%, the gap is even wider.

The Real Conversation

When I present this to practice leaders, the reaction is almost always the same. They don't dispute the math. They're not surprised that referral conversion is low. What surprises them is the dollar value of the gap.
And that's the real unlock in this conversation. Not convincing anyone they have a referral problem, but helping them see it as a revenue problem with a quantifiable number attached to it.
For PE operating partners evaluating portfolio practice performance, this is the kind of structural improvement that moves EBITDA without requiring top-line growth assumptions. For practice COOs, it's a way to capture revenue that's already in the building, just not on the schedule.
The referral pipeline is the most under-measured, under-optimized revenue source in specialty healthcare. And the practices that figure this out first aren't just going to be more efficient. They're going to be structurally more profitable.

Rick Scorzetti is Chief Commercial Officer at Parakeet Health, an AI-powered patient access platform for large specialty practices. His career spans 25 years of healthcare technology sales leadership, including senior commercial roles at Epocrates (IPO, acquired by athenahealth), WebMD/Medscape (acquired by KKR), and Jumo Health (PE exit to Falfurrias Capital Partners). Connect with Rick on LinkedIn.