If you run a practice with one to three providers, you have probably had this conversation with yourself at 9pm while staring at an aging report. Do I hire someone to sit in the office and handle billing, or do I hand it off to a billing company? Lately there is a third option too: AI-assisted billing with a human reviewing the work.
I spent ten years doing this job by hand before I started building AI agents to automate it, so I have sat on every side of this decision. There is no universally right answer. There is only what fits your volume, your risk tolerance, and how much you want billing to occupy your headspace.
What You Are Actually Deciding
Billing, sometimes called the revenue cycle (the whole process of getting paid, from scheduling a patient to posting the payment), involves a lot of moving parts: entering charges, submitting claims (the bills sent to insurance), following up on unpaid claims, fighting denials (when insurance refuses to pay), and posting payments correctly.
The question is not really "who types the claims." It is who owns the follow-up when a claim gets denied, who catches errors before they become a pattern, and who is accountable when collections dip. Keep that in mind as you weigh the three paths below.
Option 1: Hiring an In-House Biller
An in-house biller works inside your office, knows your patients, and can walk down the hall to ask a provider a documentation question. That closeness is real value, especially for a small practice where every claim matters.
The tradeoffs are also real:
- One point of failure. If your biller is out sick, on vacation, or leaves the job, claims can stall for weeks.
- Limited specialty exposure. One person has only seen the denial patterns from your practice. A billing company or a well-trained AI agent sees patterns across many practices.
- Software and training costs. Beyond wages, you are covering clearinghouse fees (the service that scrubs and routes your claims to payers), practice management software, and ongoing training as payer rules change.
If you go this route, it helps to know exactly which denial reasons are costing you the most. We wrote about the most common ones here: Why Are My Insurance Claims Being Denied? 5 Real Causes.
Option 2: Outsourcing to a Billing Company
A billing company gives you a team instead of a person. Someone is always working your claims, even when one team member is out. Good billing companies also bring cross-practice knowledge, they have likely seen your specialty's denial quirks before.
The tradeoffs:
- Less daily visibility. You are trusting a team you may only talk to once a month unless you push for regular reporting.
- Variable attention. A billing company managing many practices may not chase every small denial as hard as someone sitting in your office would.
- Contract terms matter. What counts as "collections," who handles patient calls, and who owns prior authorization (the insurance approval needed before some services) should all be spelled out before you sign anything.
If prior authorization delays are part of what is dragging down your cash flow no matter who does your billing, this checklist is worth a read: How to Speed Up Prior Authorization: A Biller's Checklist.
Option 3: AI-Assisted Billing with Human Review
This is the newer path, and it is worth understanding even if you are not ready for it yet. AI agents can be built to handle repetitive, rules-based billing tasks such as checking claims for errors before submission, tracking down the status of unpaid claims, and flagging denials for a person to review and appeal.
The point is not to remove humans from billing. It is to remove the repetitive, error-prone parts so the human on your team, whether that is you, an in-house biller, or a billing company's staff, spends their time on judgment calls instead of data entry.
This only works well when a real person still reviews exceptions and signs off on anything unusual. AI agents are genuinely good at consistency and speed on defined tasks, and they are not yet a replacement for human judgment on complicated appeals or unusual payer situations. If you are curious how far this technology has actually come versus the hype, we broke it down here: Can AI Agents Really Handle Prior Authorization Yet?.
Comparing the Cost Categories
I am not going to hand you a dollar figure, because real numbers depend on your specialty, your claim volume, and your region. But it helps to think in categories.
- In-house cost structure: salary, payroll taxes, benefits, paid time off, software licenses, clearinghouse fees, and training time. These costs stay fairly fixed whether your collections go up or down.
- Outsourced cost structure: typically a percentage of what the billing company actually collects for you, not a percentage of what they bill. This means the cost scales with your revenue, up or down.
- Hybrid or AI-assisted structure: often a flat fee or lower percentage covering the automated work, plus the cost of the human reviewer's time, whether that is your existing staff or a smaller outsourced team.
As a purely illustrative example (not a real quote): say your practice collects 100,000 dollars in a month. An in-house biller's costs stay roughly the same whether collections are 80,000 or 120,000 that month. An outsourced or hybrid percentage-based fee moves with that number. Neither structure is automatically cheaper, it depends on your volume and consistency.
For a deeper walk-through of what "fair" pricing actually looks like across these models, see Medical Billing Cost for Small Practice: What's Fair to Pay?.
How to Decide for a 1-3 Provider Practice
A few honest questions to ask yourself:
- Do I have the volume to justify a full-time salary, benefits, and software stack for one person?
- Do I trust myself to manage and audit a billing company's performance, or would I rather trust a team I can call directly?
- Am I comfortable with a hybrid model where AI agents handle the repetitive checking and follow-up, and a human handles the judgment calls?
- How much does a slow or denied claim actually cost me right now? If you have not calculated this, it is worth doing before you decide anything: What Does a Denied Claim Really Cost Your Practice?.
There is no wrong answer here, only a decision that fits where your practice is today. Small practices that outgrow one model often move to another as they add providers or change specialties.
If you want an outside, no-pressure look at where your current billing setup is leaking money, whether it is in-house, outsourced, or something in between, our free Billing Health Check at audit.getautomatedrcm.com is a reasonable place to start. It is meant to give you information, not a sales pitch.
Frequently Asked Questions
Is it cheaper to hire an in-house medical biller or use a billing company?
It depends on your claim volume and consistency. In-house billing has fixed costs like salary, benefits, and software that stay steady regardless of collections, while outsourced billing companies typically charge a percentage of what they collect, so the cost moves up and down with your revenue. Neither option is automatically cheaper for every practice, so it helps to run the numbers for your own specialty and volume.
What is AI-assisted medical billing?
AI-assisted billing uses software agents to handle repetitive tasks like checking claims for errors before submission and following up on unpaid claims, while a human still reviews exceptions and denials. It is not meant to replace billing staff entirely, but to reduce the manual, repetitive workload so people can focus on judgment calls and appeals.
How many patients or providers justify hiring an in-house biller?
There is no fixed number that applies to every specialty, but many small practices with one to three providers find that claim volume does not fully justify a dedicated full-time salary, benefits, and software stack. It often makes sense to compare the fixed cost of an in-house hire against a percentage-based outsourced or hybrid model based on your actual monthly claim volume.
Can I switch from in-house billing to an outsourced billing company later?
Yes, many practices start with one model and switch as they grow or as their needs change. The key is keeping clean records, documented processes, and a clear picture of your denial patterns so the transition does not create a gap in claim follow-up.