What medical billing really costs in 2026: percentage of collections vs. in-house

The only two ways billing is ever priced

Every medical billing arrangement in the United States comes down to one of two structures: you employ the people who do the work, or you pay someone a percentage of what they collect for you. Everything else — per-claim pricing, hybrid retainers, offshore hourly desks — is a variation on those two.

The reason practices get this decision wrong is that they compare the wrong numbers. They compare a biller's salary against a percentage fee, decide the salary is cheaper, and stop there. The salary is not the cost of in-house billing. The cost of in-house billing is the salary, plus the employer taxes, plus the benefits, plus the clearinghouse, plus the coding support, plus the vacation and sick coverage that nobody plans for, plus the receivables that age out while the one person who understands your payers is away.

Our published rate is 5–7% of what we actually collect, with a floor of $950 per month. The floor is not an extra charge — you pay the percentage or the floor, whichever is higher. A single-specialty practice sits at the bottom of that band; multi-provider groups and behavioral health, where payer rules and authorization requirements are heavier, sit higher.

What in-house billing actually costs in 2026

Build the number honestly. Start with the fully loaded cost of the billing staff you employ: base pay, payroll taxes, workers' compensation, health contribution, retirement match, paid time off. Then add the software and clearinghouse subscriptions, the coding reference tools, the credentialing work, the payer portal seats, and the share of your practice manager's week that goes to billing escalations rather than operations.

Then add the two costs nobody puts in the spreadsheet. The first is coverage risk: when one person owns the claims queue and that person is out for three weeks, the aging report does not pause. The second is the cost of the work that is never done — the small-balance denials nobody appeals, the secondary claims nobody drops, the credit balances nobody resolves.

When practices run this properly, the loaded cost of a competent in-house billing function lands well above the salary line they started with. That is the number to compare against a percentage fee, not the salary.

Why a percentage aligns the incentive and a flat fee does not

A percentage of collections means the billing partner is paid only when money reaches your bank account. If a claim is denied and never worked, nobody gets paid — and the partner loses more than you do in proportional terms, because that denial is their revenue too.

A flat monthly fee, per-claim pricing, or an hourly desk does not carry that alignment. A per-claim vendor is paid for submitting the claim, not for the outcome of the claim. That is exactly the wrong incentive for the part of the revenue cycle that determines whether a practice is paid: the follow-up.

The one honest criticism of percentage pricing is that a practice with very high collections pays a large absolute fee. That is a fair point and it is why the band is tiered by complexity rather than flat, and why the arrangement is cancellable with 30 days' notice. If the arithmetic stops working for you, you leave.

Running the comparison for your own practice

Take your last twelve months of net collections and multiply by the percentage band. Compare that against the loaded in-house figure you built above. Then adjust for the thing that usually decides it: the difference in what gets collected.

A practice moving from an under-resourced in-house desk to a managed revenue cycle usually sees collections move, not just costs. Denial rework that was never happening starts happening. Claims stop sitting past timely-filing deadlines. Secondary billing goes out. The comparison is not cost against cost — it is cost against cost, adjusted for net collection rate.

If your net collection rate is already above the mid-nineties, your denial rate is low, and your aging over 90 days is thin, in-house is probably working and you should keep it. If you cannot produce those three numbers on request, that is itself the finding.

What should stay inside the practice either way

Outsourcing execution does not outsource accountability. Fee schedules, payer contract decisions, financial and collection policy, charity care, refund approval, and access governance stay with the practice. A partner works defined queues against defined targets; the practice owns the policy those queues enforce.

Before granting access, require a written scope, a business associate agreement, role-based access, auditability, breach procedures, data-return obligations on exit, and named service levels. Our exit terms are published for exactly this reason: 30 days' notice, and your file and logins handed back within five business days.

A practical way to test it without committing

Start bounded. One location, one payer group, or one aging segment. Agree the baseline before anything moves — clean-claim rate, denial rate, days in A/R, percentage over 90 days, net collection rate — then measure the same five numbers 90 days later against the same definitions.

A free 90-day A/R and denial audit is the cheapest version of this test: we read your actual denial and aging data and tell you what is recoverable, and you keep the findings whether or not you engage us. If the answer is that your in-house team is doing fine, that is a useful thing to know with evidence behind it.