And how to tell whether yours is earning it.
Somebody asked this in a private-practice group recently: "What is typical?"
Twenty-two people replied. Six of them mentioned what they pay. Nobody answered the question.
That's not anyone's fault. There's no published benchmark for this. Billing companies don't post their rates side by side, and most practice owners have exactly one data point: their own. So the question gets asked, the thread fills up with recommendations, and everybody goes back to paying whatever they agreed to years ago.
Here's my attempt at an answer, plus the three numbers that will tell you more than any rate comparison.
This is based on the information available to me as of August 2026: rates that practice owners reported themselves in mental health practice-owner groups, alongside standard revenue-cycle benchmarks. It's a small sample of self-reported figures, not an industry survey. If your experience differs, I'd genuinely like to hear it. That's what makes the next version better.
| What they pay | What it reportedly covers |
|---|---|
| 5% of collections | Claim submission plus benefits verification and some added tasks |
| 7–8% of collections | Claim submission; added services vary and are often unclear |
| 10% of collections | Claim submission; in the case I saw, no appeals work |
| Hourly (~$40–45/hr) | A few hours a week, scoped to actual work performed |
| Per-claim flat fee | Submission priced per claim rather than as a share of revenue |
The pattern that surprised me: in this sample, the cheaper arrangements got the better reviews. The owners paying around 5% volunteered that they were happy. The owners paying 8% and 10% were the ones describing claims that went unappealed, and in one case claims that were never submitted at all, discovered by their own clients.
I want to be careful here, because it's a handful of practices and the scopes aren't identical. A 5% arrangement that includes benefits checks isn't the same product as an 8% arrangement that's submission-only. But it points at something worth taking seriously: a higher percentage isn't buying better performance, and it may be buying worse.
The most useful explanation I've heard came from an owner in one of those threads, and it reframed the whole thing: those percentages were set when billers mailed paper claims.
That was genuinely laborious work. Someone assembled, printed, and posted every claim, then chased paper remittances back through the mail. A percentage of collections was a fair way to price it.
Electronic submission collapsed that labor. In most modern practice-management systems, submitting a claim is a click once the note is signed, and posting a remittance is close to automatic. The work got dramatically smaller. The percentage mostly didn't move.
That's not an accusation against your biller. Rates drift because nobody brings them up. It's just a reason to bring it up. Two things worth knowing before you do:
Comparing rates only tells you what you're paying. It says nothing about what you're getting, and that's the part that costs real money. These are the three I'd look at, and all three can be pulled from most practice-management systems.
What it is: Of the money you were actually entitled to collect, how much did you collect?
How to figure it: Take what you were paid. Divide it by your charges minus the contractual write-downs your insurance contracts require. Those write-downs aren't losses, you agreed to them, so leaving them in makes any practice look terrible.
What I'd look for: A well-run billing operation lands around 95% or better. Consistently in the 80s usually means denials are going unappealed.
Why it beats the rate: If you're paying 8% and collecting 88% of what you're owed, your real cost is closer to 20% of your revenue — the fee plus the money nobody chased. A 5% arrangement collecting 96% is dramatically cheaper, even though the sticker prices look close.
What it is: On average, how long between doing the work and having the money.
How to figure it: Divide your current outstanding receivable balance by your average daily charges.
What I'd look for: Under 40 days is reasonable for behavioral health; under 30 is good. If it's climbing month over month, claims are aging somewhere and nobody is working the pile.
What it is: How many claims are sitting unsubmitted more than a few days after the session.
How to figure it: Most systems show claims by status. Count anything unsubmitted five or more days past the service date.
What I'd look for: Close to zero. This is the simplest of the three and the most likely to surface a real problem, because an unsubmitted claim isn't a slow payment. It's revenue that will never arrive, and it ages toward a filing deadline while it sits. It's also the failure that tends to get discovered by clients rather than owners, which is a bad way to find out.
Appeal rate. Of the claims that got denied, how many were appealed? Denials are normal, and appeals are where the money comes back. If the answer is "almost none," you're paying for claim submission rather than revenue cycle management, whatever the agreement calls it.
The most useful thing I've read on this came from an owner who had replaced two billers in two years. She caught both problems the same way: she ran reports regularly and asked questions. Her observation was that the billers who didn't work out were the ones who didn't like being asked.
That's a good filter. Questions that tend to produce informative answers:
That fourth one matters more than it looks. If your billing data lives only in your biller's system, you can't measure their performance, because they're holding the scoreboard. Getting read access to your own claim status inside your own system is worth asking for on its own.
And a framing I'd borrow from that same owner: she described contracting out as a partnership. Asking for numbers isn't an accusation. A biller who's doing well should welcome the question, because the numbers are their case for the fee.
This comes up constantly, and it usually gets debated as a difficulty question: is billing hard? Difficulty is the wrong variable. I've watched two owners with nearly identical caseloads describe the same task as "45 seconds, super easy" and "a nightmare," which tells you difficulty depends entirely on the person.
The variable that decides it is what the hours displace.
Roughly: work out what the fee costs you annually, then estimate the hours you'd actually spend. Not just submitting claims, but posting payments, working denials, appealing, and following up on aging. Divide the savings by the hours. That's what you'd be paying yourself to do billing.
Then compare it to what an hour of your clinical time earns.
If you have a waitlist, self-billing usually loses; every hour on claims is an hour not seeing someone. If your schedule has holes you can't fill anyway, it can win comfortably. And there's a third option that gets overlooked: hiring billing help hourly instead of as a percentage of collections. Several owners have landed there deliberately. It takes the work off your calendar without tying the cost to your revenue.
There's no universally right answer. There's a right answer for your practice, and it's arithmetic, not temperament.
Being straight about the limits:
Same content, formatted to print, so you can take it into the conversation with your biller instead of trying to remember which three numbers to ask for.
Free, no strings, no call required. Share it with anyone it would help.
I don't file claims. I read the numbers and tell you what they say: which of your vendor fees are percentages, when they were last priced, what you're actually collecting against what you're owed, and what a change would be worth in dollars.
Running those three numbers against your own claims data is CFO work, and it's what a monthly engagement is for. What's free is working out whether you need one — a thirty-minute checkup, no pitch at the end, and if the answer is "not yet," you'll hear that instead.
Book a free checkup See the three levels