Questions people ask
The ones that come up in almost every conversation, answered before the conversation.
Why buy this instead of parsing the insurer files myself?
You can. The files are public and some teams should. What you are paying for is the work between a published file and a number you can use: parsing terabytes of nested files, converting units so drugs are comparable, matching provider references to real providers, joining benchmarks, and doing it again every quarter when insurers change their formats without telling anyone. If you have engineers with spare time and no deadline, build it. The cost usually becomes clear after the first quarterly refresh.
How is this different from buying raw files from a reseller?
A reseller hands you the same files in a different container. This has been converted to common units, matched to real providers, compared against Medicare and the federal drug benchmark, summarized into ranges with a quality score, and joined to coverage rules. A raw file gives you a number with nothing to compare it against. This gives you the number and the benchmark next to it.
What does it cost?
The Infusion Rate Benchmark is $18,000 a year at a published price. Broader rate coverage runs $45,000 to $75,000 depending on how much you need. Policy Intelligence is $60,000 and the combined Coverage × Rate dataset is $150,000. One-off questions are $2,500 to $7,500. Everything is annual with unlimited internal users.
How often is it updated?
Every quarter, on the insurer publishing cycle. Each delivery includes a manifest with row counts, coverage and the schema version, and every column change is recorded in the changelog.
What formats do I get?
Parquet and CSV, split by quarter and state. If you need it delivered into a particular cloud bucket or warehouse, that is a delivery detail rather than a different product — tell us where it should land.
Is there any patient data in it?
No. Every source is a public document that an insurer or the federal government was required to publish. There is no protected health information in the pipeline because none ever goes in, which is why there is no business associate agreement to negotiate.
Can I put it inside a product I sell?
Yes, under an embedding license. It is priced separately because it changes the economics for both sides. Tell us what you are building and we will quote it.
Can I resell it?
Not as a dataset in its own right. Analysis, models and published findings with credit are all fine. Reselling the files is not.
Do you have SOC 2?
Not yet, and we would rather tell you than let you assume. It matters less here than usual, because there is no protected data in the product. If your process requires it, tell us early and we will tell you exactly where we are.
How many insurers, really?
Eighteen are loaded and queryable. Thirty-seven have been catalogued, which is the parsing inventory, not the product. Six groups are deliberately not in the build and they are named on the coverage page.
Can it tell me what one specific practice is paid?
Not in these datasets. They are market-level: rates summarized by insurer, state, code and setting. Named-provider detail is a different level of resolution and is in development. If that is what you need, say so, because it changes which conversation we should be having.
What if the data is wrong?
Tell us and we will check it. If it is wrong we publish the correction in the changelog against the quarter it affects rather than quietly restating it. Every coverage rule links to the insurer document it came from so you can check us directly.
Still deciding
The fastest way to answer a question about fit is to load the actual file. It downloads directly.