How we use contributed data

Only what you type into a contribution form: specialty, subspecialty, career stage, state, metro area, employer type, base salary, total compensation, $/wRVU rate and threshold, signing bonus, bonus structure, clinical FTE, years in practice, and the benefit fields on the form.

We do not scrape your employer, look you up in a compensation database, or infer your pay from anything else you do here.

• Showing you where your figures sit against published Bureau of Labor Statistics data for your specialty.
• Aggregate benchmarks shown to other physicians — but only once at least 30 physicians have contributed to the same specialty, location, and practice setting. Below that, your figures are stored and no benchmark is shown to anyone.
• Understanding which comparisons physicians actually look for, so we build the useful ones.

A “75th percentile” calculated from eight submissions is a fabricated number wearing a statistic’s clothing. So there is a hard floor: no figure of any kind — not a median, not a rounded number, not a range, not a hedge like “around” — is displayed for a cell with fewer than 30 contributors. Percentiles are not computed below the floor at all, so there is no internal version of a number we decline to show you.

Where a benchmark is shown, the number of contributors is shown next to it, never hidden behind a hover.

NPI-verified and unverified submissions are counted and displayed separately, and never averaged into one figure. Pooling them would erase the only quality signal the data has.

• We never sell it, license it, or share it with any employer, recruiter, staffing agency, hospital system, insurer, or data broker.
• We will never build an employer-facing product on it. Physicians contribute to find out whether they are underpaid; an employer would pay to learn the lowest offer a physician would accept. Same data, opposite sides. It is foreclosed by the consent text itself, so reversing it would require asking every contributor again rather than editing a policy.
• We never display your individual figures to anyone. Aggregates only, and only above the floor.
• We never use it to market a financial product to you. We do not sell any — no ads, no commissions, no referral fees.

Many physician employment agreements make the terms themselves confidential — compensation, bonus structure, non-compete language. Read your own agreement’s confidentiality section before you enter contract terms here. We cannot tell you what your agreement permits, and this is a real risk to you rather than a formality. If you are unsure, entering nothing is a reasonable choice, and the benchmark works without you.

Everything you contributed is at Settings → Your contributed data. Two separate actions, because they mean different things:

• Withdraw — the record stops counting in every benchmark immediately. It stays yours to view and export, and you can put it back.
• Erase — the record is deleted. It will not appear in your own export afterwards, and it cannot be restored.

“Immediately” is literal. Withdrawal is applied in the same database query that builds every benchmark, not by a nightly job — so there is no window in which you have withdrawn and a benchmark still contains you. Deleting your account removes your contributions from benchmarks by the same mechanism.

Each of these is a rule in code rather than an intention, and a reader who wants to check has something to point at:

• The 30-contributor floor is one guard function every display path calls, with a test that fails our build if a figure would render below it.
• The withdrawal, account-deletion, and internal-test-account exclusions are part of the query that builds an aggregate, not filters a caller has to remember.
• Consent is versioned, and a contribution records the version you were actually shown — so revising the wording later cannot re-govern an agreement you already made.
• Your compensation records are append-only: a new figure is a new record with its own effective date, so we cannot silently overwrite your history.