Frequently asked questions

An MRF is a large data file that health plans are required to publish under the federal Transparency in Coverage rule. It lists the negotiated rates a payer has agreed to for covered services, tied to provider identifiers and billing codes. The files are standardized enough to process at scale but are built for machines, not people — a single payer's file can be enormous, and the raw contents are far from a clean price list.

No. An MRF rate is a signal, not a settled payment. What a facility is actually paid depends on the revenue codes and modifiers on the claim, the structure of the contract (per diem, case rate, bundled, or percent-of-charge), which entity actually holds the contract, and whether the listed rate is even active. Two facilities can show the same number in the file and be paid very differently. A raw lookup that ignores this context misleads.

Actual negotiated rates are the real rates a payer has on file, resolved to a specific facility. Modeled or estimated rates are a formula's guess at what a rate should be. TierBench works only from actual rates — never estimates — and layers a confidence signal on top so you know how much weight each rate can bear. Real data, made trustworthy, is a fundamentally different starting point than a projection.

From payers' own machine-readable files, published under the CMS Transparency in Coverage rule. These are public, payer-published negotiated rates. We work from that public source across the country — the value we add is in resolving, verifying, and scoring it, not in the raw files themselves.

Every rate carries real-world nuance that affects how much you can trust it. Confidence scoring attaches a signal to each rate based on how it appears in the source — for example, whether it comes from a single facility's own schedule or a broader multi-group arrangement — and filters or flags what can't be confirmed. It tells you not just the number, but how solid it is. It is not a promise of a specific payment.

Our rates reflect 2026 payer data. Recency matters: MRF data circulating in the market is often stale, and a rate from years ago can misrepresent where a payer stands today. We re-ingest sources on a forward-only schedule and flag data we can't confirm as current rather than present it as fresh.

MRFs are machine-generated and full of noise: terminated contracts, duplicate entries, and legacy rates that no longer reflect anything active. We call these ghost rates. A number can be real in the file and dead in practice. Separating live rates from ghosts is part of why raw MRF data misleads on its own, and part of what our processing is built to handle.

Benchmarking places your rates against the real negotiated rates of comparable facilities, so you can see where you stand instead of guessing. For a behavioral health facility, that turns a payer conversation from 'we'd like more' into 'here is where our rate sits relative to the market' — a position grounded in data rather than hope.

Because the code is only part of the story. The same service can be paid differently depending on contract structure (per diem versus case rate versus bundled versus percent-of-charge), which entity holds the contract, whether the payer's behavioral health is carved out to a specialty network, and the facility's own negotiated position. The billing code describes the service; it does not describe the deal.

Your percentile position tells you where a given rate falls within the distribution of comparable rates — for instance, whether you sit near the middle of the market or toward the bottom. It reframes the question from 'is this rate good?' to 'where does this rate stand relative to my peers?' — which is the question that actually informs a negotiation.

You can download the files; everyone can. The gap is everything after the download: resolving rates to the real facility, untangling carve-outs, reading contract structure, filtering ghost rates, and scoring confidence. TierBench delivers intelligence you can act on, not a raw file you still have to make sense of. We sell the difference between a number and a position you can defend.

Yes — that's what Peerbench is built for. It shows how a facility's rates sit against the distribution of comparable facilities: medians, percentiles, and where you fall within the peer set. On public pages we describe the capability rather than show specific figures, but inside the product you see the real peer comparison.

Behavioral health is carved out to specialty networks more often than most of medicine, contracts vary widely by level of care, and provider identity is genuinely messy — the same organization can appear under different names and tax IDs. Each of these breaks a naive lookup. Handling them well is exactly the work that separates usable intelligence from a raw file.

Our deepest coverage comes from national behavioral health books — Carelon (Anthem/Elevance's behavioral health network) and Optum (UnitedHealth's), the two largest behavioral health networks — plus the in-house books of major commercial carriers like Aetna and Cigna, backed by regional facility coverage in select states. The Payer Coverage page shows the current, verified roster; we keep it as the single source of truth so it never drifts from what we've actually ingested.

Many payers don't administer behavioral health themselves — they carve it out to a specialty network. Anthem/Elevance's behavioral health runs largely through Carelon; UnitedHealth's runs through Optum. The name on the member's card often isn't the entity that actually holds the behavioral health contract. If you read a payer's file and stop at the brand on the card, you can miss the book that actually governs your rates.

Both. We have national reach through the major behavioral health networks, and deep facility-level coverage concentrated in select regional markets. We frame coverage honestly: broad where the national books reach, and especially deep in the regions where we've fully captured facility data.

Facility-level rates are what a freestanding behavioral health facility negotiates for services like detox, residential, partial hospitalization, and intensive outpatient — priced at the facility, not the clinician. Individual provider rates are professional fees for a single practitioner. Anyone can aggregate professional rates; facility-level data is the harder, more valuable signal, and it's where we go deep.

MRF data quality and availability vary a lot. Some files are rich and usable; others are structured in ways that make facility rates hard to trust, or are dominated by rate types that don't translate into a clean facility benchmark. We'd rather cover a payer well than list it for the sake of a longer roster, so we add coverage as we can stand behind it.

We ingest on a forward-only schedule, adding new monthly data vintages while leaving prior months immutable. Coverage grows as we ingest and verify additional payer books, and we re-verify existing sources rather than assume they stay current.

Three ways. Self-serve: explore the data directly in the platform. Custom reports: we produce a negotiation-ready analysis for your facility and payers. Consulting: behavioral health contracting expertise paired with the data, working alongside your team. You can start with any of them.

No. We don't negotiate for you — we put you in the position to negotiate, and we work beside you. You get the actual rates, the peer context, and the strategy, but you lead your own payer conversations. We're never at the table in your place.

Behavioral health facilities and groups, management services organizations, private equity and investors evaluating behavioral health assets, and the consultants and revenue-cycle firms who serve them. If your work depends on knowing what payers actually pay for behavioral health, the data is built for you.

No. The self-serve platform is built for operators and analysts, not engineers — you look up rates, compare against peers, and read percentile position without writing code or wrangling files. If you'd rather not touch the data at all, custom reports and consulting deliver the answer to you.

The self-serve platform puts the data in your hands to explore on your own schedule. Custom reports are a done-for-you deliverable: we assemble the relevant rates, peer context, and framing into a negotiation-ready analysis for your specific facility and payers. Many teams use both — the platform for ongoing questions, a report when a specific negotiation is coming up.

Ratebench and Peerbench are live today: Ratebench for looking up actual negotiated rates, Peerbench for comparing your rates against the peer distribution. Netbench and Vida are in active development — Vida is a conversational way to query the data that will never estimate or generate rates. We label what's live as live and what's coming as coming; we don't dress up in-development tools as shipped.

It replaces assertion with evidence. Instead of asking for more because costs are up, you can show where your rate sits relative to comparable facilities and make a specific, grounded case. Intelligence doesn't guarantee an outcome, but it changes the conversation from opinion to data — and it prepares you for how a payer will push back.

A defensible ask is anchored in real, comparable data: it references where your rate stands in the market, accounts for how your contract is actually structured, and holds up when a payer who knows their own numbers scrutinizes it. The opposite is an ask built on a single figure pulled out of context, which a payer can dismantle in a sentence. The goal is a position you can stand behind, not a bigger number pulled from the air.

In general, the strongest moments are ahead of a contract's renewal window — and, crucially, when you can bring new evidence to the table. A payer is far more receptive when you arrive with a clear, data-backed read on where your rates actually stand versus the market than when you simply ask for more. That evidence is what turns a renewal conversation from a request into a case. The specifics — notice periods, renewal mechanics — depend on your contracts and payer relationships, so treat this as a starting principle rather than a fixed rule.

Better data changes what you can credibly argue, which is a real advantage — but we won't promise a specific increase. Rates move based on many factors, and anyone quoting a guaranteed number is selling something. What we can say honestly is that walking in with actual, defensible peer data puts you in a materially stronger position than negotiating blind.

TierBench is built to work with you, not to replace anyone. Your team or consultant brings relationship and negotiation expertise; we bring the actual rates, peer context, and confidence scoring underneath it. The data strengthens whoever is at the table — and our own consulting option pairs the two directly when you want it.

Behavioral health care spans a continuum: medically managed detox and withdrawal, inpatient psychiatric care, residential treatment, partial hospitalization (PHP), intensive outpatient (IOP), and standard outpatient care. Each level is a distinct service with its own contracting and reimbursement patterns, which is part of why benchmarking has to be level-of-care aware.

Facility contracts are usually structured as a per diem (a flat amount per day of care), a case rate (a single amount for an episode), a bundled arrangement, a fee-for-service schedule, or a percent-of-charge. The same service can be reimbursed under any of these structures, which is why two facilities can look identical on a billing code and be paid on completely different terms.

Revenue codes describe where and how a service was delivered — the facility context around the procedure code. For behavioral health facilities they carry real weight, because the same procedure paired with different revenue codes can price differently. Reading a rate without the revenue-code context is reading one word of a sentence.

Three reasons stack up: behavioral health is carved out to specialty networks more often, provider identity is unusually messy (the same organization under different names and tax IDs), and reimbursement varies sharply by level of care. Each one breaks a naive comparison, so honest behavioral health benchmarking takes more resolution work than medical.

A per diem pays a flat amount for each day a patient is in care, so total reimbursement scales with length of stay. A case rate pays a single amount for the whole episode regardless of length. They can produce very different economics for the same care, which is why comparing a per diem against a case rate as if they were the same number is a mistake.

When a payer carves behavioral health out to a specialty network, the real contract — and the real rate — lives with that network, not the payer whose name is on the card. Carelon and Optum are the clearest examples. So a facility's actual reimbursement can be governed by an entity the member never sees, and benchmarking has to follow the rate to where it actually lives.