This is the story of an ABA practice that recovered money it had already written off — claims billed through a previous platform that were never fully collected — and, in working out how, rethought the single largest controllable line in its budget. We've kept them anonymous, but the trap they were in is one of the most common (and most expensive) in this industry.
How the "managed billing" deal actually works
Their prior setup paired the software with a revenue-cycle-management (RCM) service priced the usual way: hand over your billing, and we'll take a percentage of everything we collect. It's pitched as alignment — "we only get paid when you get paid" — and for a stretched owner who never wanted to become a billing expert, outsourcing the whole headache feels like relief.
Then you do the arithmetic. RCM services in ABA commonly charge somewhere in the range of 5–8% of collections. Run that against a practice collecting, say, $2,000,000 a year: at 6%, that's $120,000 — every year, forever, growing as you grow. For that price you could staff a billing function several times over. And you're paying it for work that is, honestly, mostly repeatable: submitting clean claims, posting payments and ERAs, tracking authorizations, flagging denials, working aging. That's not artisanal labor — it's a workflow problem software is supposed to solve.
The part that really stung: the mess was the software's
Here's what tipped the owner from annoyed to done. A good chunk of the billing "mess" the RCM service was being paid to manage had been created by the very platform collecting the percentage — claims that went out with the wrong units or a missing modifier, authorizations that quietly lapsed because nothing flagged them, data that didn't flow cleanly from the session to the claim. Paying a percentage of your revenue to have someone clean up problems the software introduced, on software you're also paying for, is a strange deal the moment you say it out loud.
And there was the other offer that gave it away: free software, if you sign up for the RCM service. Software is never free — it's just repriced as a slice of your revenue. A vendor willing to give the platform away to get a percentage of your collections has told you exactly where the money is.
The thing a percentage can't buy: the last, hardest dollar
This is the honest core of it. A managed-billing percentage buys you someone to do the work — it does not guarantee you collect everything. ABA billing leaks revenue in specific, knowable places: denials for authorization or medical-necessity, unit and modifier mismatches, timely-filing windows that close, secondary claims that never go out, small balances quietly written off. Every one of those is recoverable — but only if a human looks at the denial, decides whether it was wrong, and works it: a corrected claim, an appeal, a call to the payer.
Nobody does that for your last 10% like the person whose money is actually on the line. An outside service paid a slice regardless gets most of its fee from the easy 90% that would have paid anyway; the hard, small, aging claims are exactly the ones a detached service has the least incentive to chase. So they get written off — and you paid a percentage for the privilege.
And there's a nuance a billing service quietly depends on you not noticing: appeals only recover claims that were denied wrongly. Plenty of denials are genuinely correct — a claim that exceeds the payer's units-per-day cap (a Medically Unlikely Edit), a session outside the active authorization, a code pairing the payer simply won't allow. No one can appeal those away. The only way to not lose that money is to never send the bad claim in the first place — and that's a software job, done before submission, not an RCM job done after the denial lands. A service paid to work denials has little reason to prevent them.
What you can't hand off: your data, your visibility, your responsibility
There's a deeper version of this, and it mirrors what owners learn about labor-law compliance: outsourcing the work doesn't outsource the responsibility. An RCM service bills from the data you hand it — it does not own the accuracy or completeness of your records. If documentation is thin or the data didn't flow cleanly from session to claim, the denial is still yours to eat. The records the whole operation runs on stay your responsibility, whoever pushes the buttons.
And the moment you hand claims to an outside service, you lose sight of them. What's stuck, what's denied, what's aging, what quietly got written off — you get a check and a summary, not the claim-by-claim truth. You can't manage what you can't see, and a percentage model has no incentive to show you the parts that aren't going well.
It compounds when clinical, scheduling, authorizations, and billing live in disconnected systems and no one is watching the whole path from intake to remittance. That isn't a cosmetic problem — the seams between systems are exactly where money leaks: the unit that didn't match the note, the secondary claim that never went out, the denial nobody connected back to its cause. End-to-end visibility isn't a nicety; its absence is the thing quietly leaving money on the table.
How Wilma helped: bring it in-house, keep the visibility and the revenue
Moving billing onto Wilma changed both the economics and the visibility:
- One record, intake to remittance. Insurance is captured at intake and carried straight through to the claim, so the practice could finally see exactly what was billed, what was paid, what was denied, and — the painful part — what the old system had quietly left on the table.
- Denials and aging surfaced, not buried. Instead of a percentage and a black box, the team could systematically go back through stalled and denied claims and rework the ones that were wrongly denied — including older claims still inside their filing and appeal windows.
- Cleaner claims out the door. Authorization expirations are flagged before they turn into denials, and pre-submission claim scrubbing (part of the AI add-on) catches gaps — wrong units, a missing modifier, a documentation hole — before a claim ships. Fewer denials happen in the first place, instead of being chased after they land.
- No revenue tax. Self-service billing means the practice keeps 100% of what it collects instead of surrendering a permanent percentage. That recovered six-figure RCM fee doesn't shrink the work — software does the work — it just stops leaving.
With that visibility into the back catalog, the team recovered money from previously billed claims the prior arrangement had never closed out — dollars they had effectively given up on.
"They wanted a slice of our revenue without guaranteeing we'd actually collect all of it. Once we could see every claim ourselves, we realized we could just… do it — and keep the percentage." — Practice owner
An honest caveat — and a fair word for RCM
To be fair, revenue-cycle management has a real place. A brand-new practice, or one scaling faster than it can hire, can genuinely benefit from billing help — as an add-on that buys room to grow. The problem was never RCM itself. It's RCM sold as a substitute for owning your data, your visibility, and your collections, priced as a permanent cut of everything you earn. As help that accelerates you, it can be worth it. As a replacement for your responsibility, it's the most expensive plan you can buy — and it still leaves the responsibility with you.
Self-service billing is not zero work — someone on your team still owns it, and a practice with genuinely no billing capacity may still want help. The point isn't that billing becomes effortless; it's that the work, and the revenue, stay yours — instead of paying a permanent percentage to a service that can't promise the one thing that actually matters: every dollar collected. (To be clear about scope: Wilma captures and stores insurance at intake and scrubs claims before they ship, so the work that drives clean collections lives in one place.)
The results
- Recovered previously billed revenue the prior platform had left uncollected
- Eliminated the percentage-of-collections fee — on a practice this size, a five-to-six-figure annual cost that compounds with growth
- Full claim-by-claim visibility, so denials get worked instead of written off
- Cleaner first-pass claims, so there are fewer denials to chase to begin with
The names are hidden. The pattern isn't: if your billing is automatable — and most of it is — handing a permanent cut of your revenue to manage it, often to clean up messes the software made, deserves a hard second look. It's the same move as bringing communication in-house: when the platform does the work, you stop paying for the gaps.
Frequently asked questions
Isn't "pay a % of collections" fair since they only get paid when I do?
It sounds aligned, but it buys you someone to do the work — not a guarantee you collect everything. A percentage of revenue, forever, for work that's largely automatable is a steep, permanent tax. And no outside service works your last, hardest denials as hard as the person whose money is actually on the line.
Does an RCM service ever actually make sense?
Yes — for a brand-new practice, or one growing faster than it can staff billing, RCM can be a sensible add-on that buys room to grow. The trouble is treating it as a permanent substitute for owning your data, your visibility, and your collections, priced as a cut of everything you earn. It doesn't take responsibility for your records or your results — that stays with you. So the real question is whether you're using it to accelerate, or to avoid owning something you can't actually hand off.
If a service is working my denials, why do claims still slip through?
Because appeals only recover claims denied in error. Some denials are genuinely correct — exceeding a payer's units-per-day cap (a Medically Unlikely Edit), a session outside the authorization, a code pairing the payer won't allow — and no one can appeal those away. The only way to keep that money is to never send the bad claim, which is front-end prevention (clean, scrubbed claims) rather than after-the-fact denial work.
Why would my old billing software create the problems an RCM service then charges to clean up?
It happens more than people admit: claims denied for documentation gaps the software let through, authorizations that quietly expired, data that didn't flow cleanly. Paying extra to have someone fix problems the platform itself introduced is a strange deal once you say it out loud.
If billing is "automatable," what work is actually left for me?
The repeatable parts — clean claim submission, payment posting, surfacing denials and aging — are workflow software should handle. What's left is judgment: looking at a denial, deciding if it was wrong, and picking up the phone to the payer. That's the part that needs a human with a stake, and it's exactly where in-house beats a detached service.
Can I really collect more doing it myself?
With full visibility into every claim — what was billed, paid, denied, and left on the table — the practice in this story went back through stalled and wrongly-denied claims and recovered revenue the old arrangement had written off. You keep what you collect instead of surrendering a percentage for the privilege.
They offered me free software if I sign up for RCM — isn't that a deal?
Free software in exchange for a slice of your revenue isn't free — it's the most expensive plan you can buy, paid as a percentage forever, with no guarantee of 100% collection. Worth asking: is a cut of every dollar you earn really cheaper than self-service billing you control?
How is Wilma's billing different from just another clearinghouse?
It's self-service billing with full visibility built into the same platform as your clinical and scheduling data — insurance captured at intake and carried through, pre-submission checks (AI add-on) to catch gaps before claims ship, and denials surfaced so you can work them. You keep the revenue and the control.