Running a Medical Billing RFP: Scope, Contract and the First 120 Days

Write the scope once and send the same document to everyone. Two vendors quoting against different assumptions produce two numbers that cannot be compared, and the gap usually shows up later as an add-on fee.

The practice facts. Specialty mix and provider count; average monthly claim volume and average charge per claim; your top five payers by volume and roughly what share each represents; your EHR/practice-management system and whether the vendor must work inside it; patient-statement volume and who handles patient calls today; and whether credentialing and payer enrolment are in or out of scope.

The functions you are buying, listed explicitly so nothing is assumed: charge entry, coding review, claim scrubbing and submission, clearinghouse management and rejection handling, denial management and appeals, payment posting, A/R follow-up with an aged-bucket policy, patient statements and patient calls, credentialing, monthly reporting, and a named account manager.

Your own starting numbers, with the definition you used for each: adjusted (net) collection rate, days in A/R, denial rate, aged A/R over 90 days, and first-pass clean claim rate. These are the baseline the transition gets measured against, and stating your definition prevents a vendor from reporting against a friendlier one.

What you want back, in writing, from every vendor: an all-in price against this exact scope with every fee itemised; the KPI targets it will commit to and the definition of each; contract term and notice period; data-export format and timeline on exit; the transition plan with dates; and where its billers and coders are located.

Before you compare the prices that come back, run each one through the effective-rate arithmetic on the best medical billing companies page — a headline percentage plus a software fee plus a statement fee plus an amortised setup fee is a different number from the one on the proposal.

Price is the term buyers negotiate and the one that matters least once the engagement is running. These are the ones to settle first.

Term and notice. A short initial commitment (60–90 days is common) followed by month-to-month with 30–60 days' notice. A 24- or 36-month term with an early-termination fee protects the vendor, not you, and it removes the only leverage you have if performance slips.

Every fee itemised, not just the percentage. Setup, implementation, software or platform access, per-claim charges, patient statements, patient calls, credentialing, appeals, minimum monthly billing and any charge for a data export. A fee that is absent from the headline rate is the one that changes your effective rate after you sign.

KPI targets with their definitions attached. Agree the metric, the definition, the data source (your practice-management system of record, not the vendor's spreadsheet), the reporting cadence, and what happens if a target is missed. Definitions are where reported performance gets quietly manufactured — a collection rate quoted against charges rather than against allowed amounts is a different and much larger number.

Turnaround commitments. Charge entry, claim submission, clearinghouse rejection handling, payment posting and patient statements each get a stated timeframe. “Promptly” is not a commitment.

Data ownership and exit. Your data is yours: put the export format, the scope of the export (patient records, claim history, payment history, open A/R), the timeline and the cost in the agreement now. A vendor that will not write that down has told you what leaving will look like. Our own guide to switching billing companies without losing revenue covers the operational half of the same problem.

Compliance documentation. A signed HIPAA Business Associate Agreement, written security policies, documented breach-notification procedures, role-based access control, evidence of workforce HIPAA training, and the date of the vendor's most recent risk assessment. Ask what independent attestation the vendor holds and treat “we are HIPAA compliant” as an unaudited self-assertion until it is evidenced — including from us: ask us what attestation we hold, what its scope is and what we can send you under NDA, and hold our answer to the same standard as everyone else's.

A professional onboarding plan is a project plan with dates on it, not a promise. Ask for it in writing before signing, and ask the vendor to tie it to your EHR and your payer mix rather than to a published norm — the timeline is decided by payer enrolment and EDI/ERA setup, which vary by payer and by state.

What the plan should contain: a discovery phase covering payer enrolments, EDI/ERA setups, fee schedules and workflow mapping; EHR/practice-management integration with test claim runs; a parallel run or a structured cutover that protects in-flight claims; training for your staff on the new touchpoints; and a named implementation lead with weekly status reporting.

Milestones worth dating in the plan: a clean handoff of in-flight A/R; the first full submission cycle; the first KPI scorecard measured against the baseline numbers you put in the RFP; a denial-root-cause report with provider-level and payer-level detail; and a formal performance review against the contracted targets.

What to watch during the run. Two things reliably go wrong in a transition and both are detectable early: claims stalling at the clearinghouse because an enrolment has not completed, and the previous vendor's aged A/R being quietly deprioritised in favour of new claims. Ask for both to be reported separately from week one — aggregate collections will hide them for a month or more.

Measure against your own pre-transition baseline, not against an industry figure. The only freely published physician-practice targets we could source are the AAFP's practice-management guidance (adjusted collection rate 95% at minimum; days in A/R below 50 at minimum, 30–40 preferable; denial rate 5%–10% as the industry average), and those are guidance without a stated population, sample or data year. MGMA's medians sit inside licensed products and are not public, and HFMA's MAP Keys publish definitions rather than target values — so a vendor attributing a target number to either is citing something that does not exist in public.

Memorialise the targets before signing, then actually read the report. The review that matters is monthly, takes twenty minutes, and asks four questions of the same numbers every time: is the adjusted collection rate moving against our own baseline; is aged A/R over 90 days growing as a share of total A/R; which denial reasons are repeating, by payer; and which of those are documentation or front-desk problems on our side rather than the vendor's.

Insist that reporting breaks down by payer and by provider rather than arriving as practice-level aggregates. An aggregate hides the single payer or the single provider that is generating the problem, which is exactly the detail a remediation plan needs.

If a target is missed, the contract should say what happens next — a written root-cause response within a stated number of days, a remediation plan, and a review date. Without that, a missed target produces a conversation instead of a change. And if you conclude the engagement is not working, the data-export terms you settled before signing are what make leaving a decision rather than a negotiation.