Benchmarks: MGMA, societies, and the samples behind the numbers

Three source types, three different samples

Every practice benchmark you'll encounter falls into one of three categories: official government statistics, professional-society member surveys, or proprietary vendor reports sold by subscription — and the category tells you more about how much to trust a number than the number itself does. Government data is compiled from mandatory or near-universal employer reporting; society surveys and vendor reports are both built from whoever chose to respond, which is a fundamentally different kind of sample.

None of these categories is inherently wrong to use. The mistake is treating a number from one category as if it carried the reliability of another — a self-reported survey median presented with the same confidence as an audited government statistic is the most common way a benchmark misleads.

Official government data: free, large-sample, narrow scope

The Bureau of Labor Statistics publishes wage distributions — median and percentile compensation by occupation, state, and metro area — for clinical and counseling psychologists 1Ref 1U.S. Bureau of Labor Statistics (2025).Occupational Employment and Wages: Clinical and Counseling Psychologists.Official government wage distribution for psychologists, used as an example of the government-data benchmark category., healthcare social workers 2Ref 2U.S. Bureau of Labor Statistics (2025).Occupational Employment and Wages: Healthcare Social Workers.Official government wage distribution for healthcare social workers, used as an example of the government-data benchmark category., and substance abuse, behavioral disorder, and mental health counselors 3Ref 3U.S. Bureau of Labor Statistics (2025).Occupational Employment and Wages: Substance Abuse, Behavioral Disorder, and Mental Health Counselors.Official government wage distribution for mental health counselors, used as an example of the government-data benchmark category., drawn from employer-reported wage surveys at government scale rather than a self-selected opt-in panel. This is the most trustworthy category of the three for what it actually measures.

Its limitation is scope, not reliability: BLS wage data tells you what employees in these roles earn, not what a solo private-pay practice's revenue per visit or overhead ratio looks like — those figures simply aren't federal wage statistics, so government data can anchor your compensation comparison but can't answer most of your other practice-metrics questions.

Federal payment and quality data: public, but not a P&L benchmark

CMS publishes some of its own landscape data — the Innovation Center's page on the alternative payment models most value-based arrangements trace back to is one example 4Ref 4Centers for Medicare & Medicaid Services (2026).CMS Innovation Center.Example of a federal source publishing payment-model landscape context, distinguished from a practice-financial benchmark. — but this kind of federal publication is context about how the system is structured, not a practice-financial benchmark you can compare your numbers against directly. Read it as background on where reimbursement is heading, not as a source for "what should my overhead be."

Quality-measure benchmarks sit in a related but distinct category: HEDIS is the measure set health plans report, and several of its measures — antidepressant medication management, follow-up after an ED visit for mental illness — reach into outpatient behavioral-health practice patterns 5Ref 5National Committee for Quality Assurance (2026).HEDIS.That HEDIS measure definitions are public even though plan-level performance against them usually isn't, distinguishing quality benchmarks from financial ones.. The measure definitions are public; how any specific plan's population actually performs against them usually isn't, since plans treat their own HEDIS results as competitive data.

Professional-society surveys: real practices, self-selected sample

Professional bodies like APA's practice organization publish practice-management, reimbursement, and compensation guidance drawn substantially from their own membership 6Ref 6APA Services, Inc. (2026).Practice — APA Services.Example of a professional-society source publishing practice-management and compensation guidance drawn from its own membership, used as the survey-derived benchmark category. — a genuinely useful source, and one with a real professional-body anchor behind it, but built from members who chose to respond to a survey rather than from a mandatory or randomized sample. Members who are doing unusually well, or unusually poorly, may be more or less likely to respond than the typical member, and that skew is rarely disclosed in the published summary.

Label any figure from this category as survey-derived when you use it, even mentally — "the society survey found X" is a meaningfully different claim than "the government reports X," and treating the two the same is how a soft number quietly becomes a hard target.

Proprietary vendor reports: paid access, larger-practice skew

The largest commercial benchmark vendors sell aggregated practice-financial data — collections per hour, overhead ratios, days in A/R, staffing ratios — by annual subscription, built from whichever practices submit data to them, typically in exchange for access to the aggregated results. That exchange mechanism tends to pull in larger group practices and health systems with dedicated staff to do the data submission, which skews the resulting averages toward organizations that don't look like a solo practice.

Overhead at micro scale in particular reads very differently from what a subscription benchmark vendor reports as "typical" overhead, precisely because a one-clinician practice has a cost structure a multi-provider group doesn't share — rent and a single EHR license behave nothing like the overhead line items in a report weighted toward group practices. Treat a vendor benchmark as a reference point from a different population, not a target.

Four questions to ask before trusting any benchmark number

Before comparing your practice against a published benchmark, check who's actually in the sample: what practice size and setting dominate it, whether the data is self-reported or independently verified, how recent the collection period is, and whether the payer mix behind the number resembles yours. A benchmark built mostly from group practices with a favorable payer mix tells you little about a solo, private-pay-heavy practice, no matter how official the publisher sounds.

• Sample size and composition — a report built from a few hundred large practices isn't the same population as a solo practice
• Self-reported vs. verified — survey and vendor data is usually the former
• Recency — reimbursement and staffing costs move fast enough that a three-year-old benchmark can mislead
• Payer mix — a benchmark from mostly-insurance practices doesn't transfer cleanly to a private-pay practice

Your own numbers, tracked consistently, beat someone else's average

External benchmarks are most useful as a sanity check, not a target to hit — a number wildly outside any published range is worth investigating, but a number that simply differs from a group-practice-weighted average often just reflects that you're a different kind of practice, not that something is wrong. Your own trend over time, measured the same way every month, is a more honest signal than a cross-sectional comparison against a population you don't belong to.

The solo dashboard is built on exactly this logic: a small set of numbers tracked consistently against your own history, so "is this normal" gets answered by your own trend line first and an external benchmark second, if at all.