Percentage of Collections vs Per-Claim Pricing

Percentage-of-collections pricing means the vendor invoices a stated percent of your net collections each month. 'Net collections' is a critical defined term: it should mean dollars actually received from payers and patients in the billing period, after refunds and write-offs, but excluding refunds of overpayments and excluding patient credits. The percentage is typically applied to all collected dollars, not just collections from claims the vendor submitted, which means the vendor is paid on aged A/R recoveries from before the relationship started — read your contract on this, because it varies. Quoted rates commonly run about 4-9%, and specialty-billing services such as mental health billing services, anesthesia and pain management tend to sit at the upper end because of higher per-claim complexity. The bottom of that range has one freely-public survey behind it — the December 2025 vendor survey recorded in our Benchmarks Report found 63% of 190 billing companies charging 7.99% of collections or less — but nothing public establishes the 9% top end, which remains an observed quote rather than a surveyed figure. The invoice is monthly, calculated on the prior month's actual receipts.

Per-claim pricing means the vendor invoices a flat fee for each claim submitted, regardless of whether the claim is paid, denied, or partially paid. Quoted rates commonly run $4-$8 per claim, with cleaner-volume vendors at the lower end and specialty/complex-claim vendors at the higher end - again an observed range of vendor quotes, not a surveyed figure. Some per-claim vendors charge a separate fee for resubmissions (so a denied-and-resubmitted claim becomes two billable submissions), which materially changes the economics. Some per-claim vendors offer a denial-recovery add-on (a separate per-denial-worked fee) but most do not, which is the structural incentive problem covered below.

Hybrid models exist: a base monthly fee plus a small percentage, a per-claim fee plus a percentage of denial recoveries on a contingency basis, or per-FTE pricing where you essentially rent a dedicated biller. These are minority structures; the % vs per-claim choice covers most contracts.

The breakeven point between the two models depends on average payment per claim and denial rate. Walk through the example math to see why.

Example 1, primary-care practice: average payment per claim is $115, monthly claim volume is 1,200, monthly net collections are $138,000. At 6% of collections, the vendor invoice is $8,280 per month. At $5 per claim, the invoice is $6,000 per month. Per-claim wins by $2,280 per month, or roughly 27%.

Example 2, mental health practice: average payment per claim is $135 (90837 hourly psychotherapy), monthly claim volume is 800, monthly net collections are $108,000. At 7% (specialty rate), the vendor invoice is $7,560. At $6 per claim, the invoice is $4,800. Per-claim wins by $2,760, or 36%.

Example 3, orthopedic surgery practice: average payment per claim is $620 (mix of E&M and surgical procedures with high RVUs), monthly claim volume is 400, monthly net collections are $248,000. At 6%, the vendor invoice is $14,880. At $6 per claim, the invoice is $2,400. Per-claim wins by $12,480, or 84%.

Example 4, pediatric practice with 12% denial rate: average payment per claim is $95, monthly submitted claim volume is 1,500 (but only 1,320 actually pay due to denials), monthly net collections are $125,400. At 6%, the vendor invoice is $7,524. At $5 per claim on 1,500 submissions plus $5 per resubmission on 180 denials, the invoice is $8,400. Percentage wins.

The pattern: per-claim pricing arithmetically beats % of collections when average claim value is moderate-to-high; % of collections wins when claim volume is high relative to revenue or when denials require substantial rework. The breakeven on average claim value (at 6% vs $6 per claim) is exactly $100 per claim; below $100, per-claim is more expensive, above $100, % of collections is more expensive.

The arithmetic favors per-claim in many practice profiles, but the incentive structure does not. This is the most underappreciated factor in pricing-model selection.

Under percentage-of-collections pricing, the vendor's revenue is a direct function of what you actually collect. A denial that does not get worked is the vendor's lost revenue too. An aged A/R balance that goes unrecovered is the vendor's loss too. This creates strong, mechanical incentive to work denials, follow up on aged A/R, appeal underpayments, and recover patient responsibility — because every dollar collected is some cents to the vendor.

Under per-claim pricing, the vendor is paid on submission, not on collection. A clean claim that pays first-pass and a claim that denies and goes unworked produce identical revenue to the vendor. The vendor's incentive is throughput: submit as many claims as quickly as possible. Denial work, A/R follow-up, and appeals are pure cost to the vendor without offsetting revenue. Some per-claim vendors offset this with internal SLAs and quality controls, but many do not, and the result is a structural exposure rather than a measured one: if nobody is paid to work a denial, denials go unworked, and the practice absorbs the loss. We found no freely-public study measuring net collections across the two pricing models at comparable practice size, so no outcome gap is quoted here — the argument is about who is funded to do the work, which is verifiable from the contract itself.

The practical implication: per-claim pricing is structurally appropriate only for practices with already-strong denial rates (under 4%) and stable A/R, where there is no material denial-work upside the vendor would need to be incentivized to capture. For practices with elevated denial rates, the apparent per-claim savings are often illusory because the vendor leaves recoverable revenue on the table. How large that gap is for a given practice is a question only its own collection data can answer; no figure is quoted for it here, because none is published.

Pricing models also allocate financial risk between practice and vendor differently. Practice owners often miss this dimension entirely.

Under percentage-of-collections, the vendor shares downside risk with the practice. A bad payer-mix month (high commercial-payer denials, slow Medicare adjudication, etc.) reduces the vendor's invoice along with the practice's revenue. A practice that loses a major payer contract sees its vendor invoice fall proportionally. The vendor is a partner in revenue performance; their cash flow tracks yours.

Under per-claim pricing, the vendor is insulated from collection risk. The vendor invoices the same per-claim fee whether collections boom or collapse. A bad month for the practice is the same revenue month for the vendor. A practice that loses a payer contract continues paying per-claim fees on whatever residual volume exists. The vendor's cash flow is decoupled from yours.

For practices with stable, predictable payer mixes and revenue, this risk-decoupling does not matter much. For practices with volatile revenue (new practice ramping up, practice in a major payer renegotiation, practice in an acquired or merging market), the % model's downside-sharing is materially valuable. For very large stable practices that view billing as pure operations, the per-claim model's predictability has real planning value.

This is also why hybrid models (small base fee + small percentage) exist: they distribute risk in a way that pure-% or pure-per-claim does not. A small base ensures the vendor cannot lose money on a slow month; a small percentage keeps incentive alignment on collections.

Once you choose a pricing model, the contract clauses around it determine whether the headline rate is the actual cost. The following are the highest-leverage clauses to negotiate.

For percentage-of-collections contracts: define net collections explicitly (which payer adjustments and patient credits are included or excluded), specify whether the percentage applies to aged-A/R recoveries from before the contract started (some vendors charge full % on legacy A/R they barely worked), specify a floor (a minimum monthly invoice that protects the vendor) only if you want to share that risk, specify a ceiling on invoice as % of revenue (rare but worth requesting on high-revenue practices), and specify how refunds and recoupments affect prior invoices (clawback or no clawback).

For per-claim contracts: define what counts as a 'submitted claim' (corrected claims and resubmissions are the most common dispute), specify whether secondary and tertiary claims count as separate submissions or are bundled with the primary, specify whether denial-rework triggers an additional fee, specify whether eligibility verification and prior authorization carry separate fees, and specify the volume tier breakpoints if pricing is volume-discounted.

For both: contract a 60-90 day termination clause without auto-renew penalty, contract data-portability (return of all PM-system credentials and historical data on termination), contract a no-poach of patients/referrers, and contract KPI-tied performance SLAs (denial rate, days-in-A/R, net collection rate) with right to renegotiate or terminate on SLA failure. The headline rate is rarely the largest cost variable; the contract architecture usually is.

Distill the cost math, incentive structure, risk allocation, and contract considerations into clear practice profiles.

Choose percentage-of-collections if: your specialty has high denial complexity (mental health, oncology, cardiology, pain management); your average collected revenue per submitted claim is below the break-even (per-claim fee ÷ percentage — $100 at $6 and 6%), so the percentage is also the cheaper invoice; your current denial rate is over 6% and you need the vendor incentivized to drive it down; your practice is in growth mode and you want vendor cost to scale with revenue rather than as a fixed line; you have meaningful aged A/R that needs recovery (over 20% of A/R aged 90+ days); or you are a new practice ramping up and unable to predict claim volume.

Choose per-claim if: your average collected revenue per submitted claim sits well above the break-even (high-RVU specialty surgery, ortho, cardiology procedures, certain dental codes), where the percentage model's arithmetic premium compounds; your current denial rate is under 4% and stable; your claim volume is predictable and you want fixed-cost predictability for budget planning; your claim volume is high enough to pull the per-claim quote to the bottom of the range (urgent care, primary care, lab) — recompute the break-even at that lower fee, because a genuinely low-dollar claim book is where the percentage becomes the cheaper invoice; or you are a very large practice (20+ providers) where per-claim economies-of-scale make the math decisive.

Consider hybrid if: you want strong denial-work incentive but are uncomfortable with pure-% on a high-revenue/high-claim-value mix. A small base + small % can capture the right incentives without paying full percentage on every collected dollar.