How we turn public payer data into rate intelligence you can trust.
The negotiated rates are public — but a raw machine-readable file is a long way from usable reimbursement truth. This is the work in between, and why you can stand behind the numbers you get.
The source: Transparency in Coverage
It all starts from the same place every payer is required to post — and the same wall every team hits when they try to use it themselves.
Legally public
Under federal Transparency in Coverage rules, payers must publish their negotiated rates in machine-readable files. The data is public — but public isn't the same as usable.
Built to resist you
A single payer's posting can run to terabytes of nested data across multipart files — far past what a spreadsheet or a quick script can handle.
A schema that fights back
The CMS file structure is sprawling and uneven, and payers forward-date, backdate, and silently revise postings — so yesterday's extract is already drifting.
The methodology, end to end
Six stages turn the raw firehose into clean, comparable intelligence. The brutal parts stay on our side; you get the result.
- 01
Extract
We pull negotiated-rate records out of enormous, deeply nested payer files at national scale — the part most teams never get past.
- 02
Normalize
Wildly inconsistent payer formats are reconciled into one comparable shape, so a rate from one payer lines up against another.
- 03
Resolve entities
The same provider appears under many NPIs, TINs, and DBAs. We resolve them to the real operating organization, so rates map to who actually delivers the care.
- 04
Detect carve-outs
The payer on the file often isn't the entity that sets the rate — behavioral health is frequently carved out to a managed vendor. We detect that and attribute the rate to the true contracting entity.
- 05
Map service lines
Billing codes are grouped into the service lines you negotiate — residential, detox, PHP, IOP, outpatient, and more — so you compare like for like (e.g. IOP, billed under H0015).
- 06
Benchmark
Clean, resolved rates become peer benchmarks — medians, percentiles, and the distribution of comparable facilities in your market.
Why a rate isn't just a rate
This is the nuance that earns the confidence framing. Two rates with the same code can mean completely different things — and acting on the wrong reading is worse than having no number at all.
The same code, priced differently
One billing code can reimburse at different amounts depending on the revenue code or modifier it's paired with — the headline code alone doesn't tell you what gets paid.
Facility vs. group vs. individual
A rate attached to a facility, a provider group, and an individual clinician are not the same number. Read the wrong one and your benchmark is off before you start.
Contract structure changes everything
Per diem, case rate, bundled, and percent-of-charge arrangements aren't comparable at face value. We account for the structure so the comparison is honest.
Ghost rates
Terminated agreements, duplicates, and legacy postings linger in the files. Treated as real, they quietly poison a benchmark — so we identify and filter them.
Confidence scoring: actual rates, made trustworthy
We don't hand you a number and walk away. TierBench scores how much to trust each rate using real signals — for instance, whether it comes from a single facility's own schedule or a broader multi-group arrangement — and filters or flags what can't be trusted: terminated, duplicate, ghost, and legacy postings.
The rates stay real — never modeled estimates. The score just tells you how firmly to lean on each one.
Illustrative concept — competitor dollar figures stay hidden by design.
Illustrative example of confidence scoring. Three sample rows, each a billing code and service line with a percentile position and a confidence badge: residential under code H0018 scored High from a clean single-facility schedule; partial hospitalization under H0035 scored Medium from a multi-group aggregated arrangement; detox under H0010 scored Low from sparse, aged data and flagged. No dollar amounts or named-payer rates are shown.
Machine speed, human judgment
The hardest engineering is already built and running — so the slow part of analysis is done before you ask. People do the rest.
The heavy lifting is already done
Ingestion, resolution, and verification run continuously across markets. When a question comes in, the clean, comparable data is already there — so turnaround is fast.
Experts at the controls
Our analysts bring deep behavioral-health domain knowledge and drive Ratebench and Peerbench as expert operators — so you get judgment and context, not just a query.
What this means for you
- Clean, comparable rate intelligence — not a pile of raw files
- Rates resolved to the real provider and the true contracting entity
- A confidence signal on every rate, with unreliable data filtered or flagged
- Benchmarks you can defend in a payer conversation
See the methodology on your rates.
Explore the data yourself, or book a 20-minute call and we'll walk through how it applies to your contracts.