Rate context you can defend in the room.

Observations from the payer's own published files

The numbers come from the machine-readable files payers publish under federal price-transparency rules — the payer's own disclosure, not a survey and not a model. A source registry keeps track of where each payer's files live, how often they change, and whether the last pull actually worked.

• Transparency page, index URL and file pattern tracked per payer
• Last checked, last successful pull, and update frequency on every source
• Ingestion and parser status, so a stale source is visible rather than silently missing
• Selective streaming filtered to your codes, your NPIs and TINs, and your markets

A Medicare anchor from public CMS sources

Every rate is shown against the Medicare amount for the same code, locality and setting, drawn from the CMS physician fee schedule. Percent of Medicare is the language payer contracting conversations are already conducted in, and it is the one benchmark both sides of the table can independently verify.

• CMS physician fee schedule anchor by code, locality and facility or non-facility setting
• Percent of Medicare on your own contract rate and on every observation
• Rates expressed by the payer as a percentage of Medicare are labeled as formulas, not converted into a dollar amount we invented

Provider matching, labeled honestly

Matching a rate in a payer file to your organization is the hard part of this data, and pretending otherwise is how a number ends up in front of a payer and falls apart. Every observation states how it was matched, and weaker matches are labeled rather than quietly averaged in.

• NPI plus TIN: matched against your roster
• Group level: provider group matched, no TIN in the file
• NPI only, approximate: no TIN available, geography or entity inferred
• Match strength is a visible column, not a hidden weighting

A data-quality label on every figure, with its reason

High, medium and low are not decorations — each label carries the sentence that explains it, and low-quality rows are excluded from the observed range rather than dragging it around. You can hide them entirely, and the range tells you how many observations it is built from.

• Reasons stated in the row: exact code and clear rate type, group-level match with inferred geography, ambiguous match with conflicting duplicate rows
• Observed range reports how many observations are usable and how many were excluded
• Low-quality observations can be hidden in one click
• Rate type and billing class shown alongside the amount — a professional component is never compared against a global rate

The provenance drawer: everything behind the number

Any figure opens into its own evidence: which file it came from, which index listed that file, when the payer published it, when we last refreshed it, which plan and product it applies to, and what scope the number covers. If a figure is going into a payer conversation, the person carrying it can see everything supporting it first.

• Source file and index, with the payer's publication date
• Last refresh, and the age of the observation in days
• Payer, plan, product, market and setting
• Export with sources, or send the code straight into a negotiation packet

Honest about coverage

Rate coverage is built market by market, payer by payer, and code by code — starting with the codes and payers our customers actually bill. Where an observation does not exist for your code, product or market, the product says so instead of substituting a national average and hoping. No naked numbers, and no implied completeness.

• Coverage is stated per code, payer and market rather than claimed globally
• Codes without usable observations are shown as such, with the Medicare anchor still available
• Nothing is displayed without a source, a date, a scope and a quality label