Is My Practice Overhead Too High? Where the Money Actually Goes
Almost every physician who asks whether their overhead is too high is really asking two questions at once: is my percentage bad, and is it bad for my specialty. The second question is the one that matters, and it is the one a single national figure cannot answer.
Before either question is worth asking, though, the number has to be built correctly — because the most common reason a practice looks like it has an overhead problem is that it counted something that does not belong in the numerator, or divided by the wrong denominator.
Quick answer
A practice at 45–65% of net collections is inside the normal band, and where you should sit inside it depends almost entirely on your specialty: primary care typically runs 45–50%, surgical and equipment-heavy specialties 60–70%. Staff salaries and benefits are the single largest line by a wide margin — on a $1.5M practice with $760,000 of overhead, staff alone is $400,000, or 53% of all overhead. The condition that changes everything: physician and advanced-practice-provider compensation is not overhead. Include it and you will fail a benchmark you actually passed.
The definition is where most practices go wrong
Overhead is the cost of operating the practice, before any provider is paid. It includes clinical and administrative staff, rent and facility costs, medical and office supplies, malpractice premiums, utilities, IT and EHR, billing, equipment and maintenance.
It excludes physician compensation, physician benefits, and owner distributions. And — this is the line that trips people — it also excludes the compensation of non-physician providers: the nurse practitioners and physician assistants who generate their own encounters. Benchmark surveys report operating cost against total provider compensation, so if you fold your NP's salary into "staff costs" you are comparing a numerator that includes provider pay against a benchmark built without it.
The size of that error is not small. Take the $1.5M practice the practice overhead calculator loads by default, with $760,000 of true overhead — 50.7%, inside the normal band for any specialty. Move a single $140,000 NP salary into the staff line and overhead becomes $900,000, or 60.0%. Nothing changed. You did not overspend, hire, or lose a contract. You crossed nine points of benchmark on a bookkeeping decision, and if you then act on it by cutting real costs, you are solving a problem you invented.
The denominator does the same trick in reverse. Benchmarks are stated against net collections — money actually received — not gross charges. Feed the same $760,000 against $2.4M of billed charges and the tool returns 31.7%, a number that would be extraordinary if it meant anything. It does not. Charges are a list price nobody pays. Use collections.
One practice, carried end to end
Here are the overhead calculator's default figures, worked through completely.
| Line | Amount | % of $1.5M collections | % of total overhead |
|---|---|---|---|
| Staff salaries and benefits | $400,000 | 26.7% | 52.6% |
| Rent and facility | $120,000 | 8.0% | 15.8% |
| Supplies and materials | $80,000 | 5.3% | 10.5% |
| Malpractice insurance | $60,000 | 4.0% | 7.9% |
| Everything else (utilities, IT, equipment) | $100,000 | 6.7% | 13.2% |
| Total overhead | $760,000 | 50.7% | 100% |
| Available for physician compensation | $740,000 | 49.3% | — |
Two cautions on that last row. It is pre-tax and pre-benefits, and it is the pool for every owner physician together, not one doctor's pay — which is why the calculator asks how many owners share it. Enter three partners and the same $740,000 reads $246,667 each.
The comparison line is specialty-matched, not national. Against the office-based primary-care range of 45–50%, this practice comes back 0.7 points above the top of the range: marginally high, worth a look at staffing. Tell the same tool it is a surgical or procedural practice and the identical 50.7% reads 9.3 points below the 60–70% range: enviable. Same practice, same arithmetic, opposite verdicts — which is precisely why a single national average is worse than no benchmark at all, and why the tool asks for your specialty before it compares anything.
Why the healthy range is not one number
The reason a range as wide as 45–65% can all be "normal" is that overhead is a ratio, and specialties differ enormously in both terms. The calculator's own component bands show where the variation lives:
| Category | Typical band, % of collections |
|---|---|
| Staff salaries and benefits | 25–30% |
| Rent and facility | 8–12% |
| Supplies and materials | 5–8% |
| Malpractice insurance | 4–6% |
| Utilities, IT, equipment maintenance | 5–10% |
| Total | 47–66% |
An office-based primary care practice sits at the bottom of nearly every one of those bands and lands around 45–50%. A surgical or procedural specialty carries more clinical staff per provider, more expensive equipment on a maintenance contract, more supply cost per case, and a higher malpractice premium — and lands at 60–70% on exactly the same arithmetic.
Medicare's own fee schedule is built on this fact: each code carries a separate practice expense RVU reflecting clinical staff time, supplies and equipment, adjusted by geographic practice cost indices, and paid at a different rate depending on whether the service is performed in a facility or in your office. CMS pays a higher practice-expense component precisely because the overhead is higher. The reader's takeaway is not that a surgical 65% is fine and a primary-care 65% is not — it is that comparing yourself to any figure that is not specialty-matched and region-matched tells you almost nothing.
The same caution applies to cost per RVU, the other ratio practices reach for. The calculator will divide your overhead by your own annual work RVUs if you enter them, and omits the figure entirely if you leave the box blank — because there is no default production level worth assuming on your behalf. The same $760,000 of overhead is $190 per wRVU at 4,000 wRVUs and $95 at 8,000, and those are two different practices carrying an identical cost base. Like the percentage, the number only means something against a specialty-matched source.
Staff is the number; everything else is rounding
At 52.6% of all overhead, staff is larger than rent, supplies and malpractice combined. Two consequences follow.
First, no meaningful overhead reduction happens without touching staffing — schedule, mix, or productivity. Renegotiating supply contracts for a 10% saving frees $8,000 and takes overhead from 50.7% to 50.1%: half a point. The same 10% from staff frees $40,000 and takes it to 48.0%, five times the effect.
Second, and more usefully: because staff cost is largely fixed within a range, the ratio moves faster when volume moves than when costs move. Divide your overhead by your average net collection per encounter to see how many visits simply pay for the doors being open. At $760,000 and a hypothetical $150 net per visit, that is 5,067 encounters before the first dollar reaches a physician — the fixed-cost arithmetic the breakeven calculator runs directly.
When overhead "rises" and nothing got more expensive
This is the most common false alarm in practice finance, and it is worth carrying through.
Hold every cost in the table exactly where it is — $760,000, not a dollar more — and let collections fall 10%, from $1.5M to $1.35M. One payer renegotiated, or a partner cut back, or the schedule thinned:
| Before | After | |
|---|---|---|
| Net collections | $1,500,000 | $1,350,000 |
| Total overhead | $760,000 | $760,000 |
| Overhead rate | 50.7% | 56.3% |
| Physician compensation pool | $740,000 | $590,000 |
Overhead "rose" 5.6 points and the physicians lost $150,000, with zero cost increase. Run both years through the practice overhead calculator side by side and the diagnosis is immediate: if the overhead dollars barely moved, you do not have a cost problem.
The choice of remedy follows from that. To get back to 50.7% by cutting alone, you would have to remove $76,000 of annual cost from a base that is half staff — in practice, about one full-time position, with the service consequences that implies. Recovering the $150,000 of collections restores the ratio without any of it. Cost discipline is the right answer when the dollars grew; volume, coding accuracy and contract terms are the answer when they did not.
Where a defensible benchmark actually comes from
If you want a number you can take to a partners' meeting, there are exactly three honest sources, and none of them is a blog post.
MGMA's cost and revenue survey is the reference most practice administrators mean when they say "benchmark". It is specialty-specific and it is genuinely good — and it is paywalled, sold to subscribers and to practices that submit their own data. Anyone quoting you a precise MGMA figure for free is either a subscriber or making it up.
Your specialty society frequently publishes a practice-cost survey to members, often at lower cost and closer to your actual case mix than a general benchmark.
Your own practice, last year. This is the one people skip and it is the most useful. A same-practice, same-definition comparison across two or three years controls for specialty, region, payer mix and your own accounting conventions all at once. Any external benchmark you buy still has to be reconciled against your definitions before it means anything; your own trailing history does not.
FAQ
Should my own salary be part of overhead?
No. Physician compensation, physician benefits and owner distributions all sit below the overhead line — overhead is what the practice costs to run before any provider is paid. The same exclusion applies to nurse practitioners and physician assistants who generate their own encounters. In the worked example above, including one $140,000 NP salary in staff costs moved reported overhead from 50.7% to 60.0% with no change in spending.
My overhead is 62%. Is that a problem?
It depends entirely on your specialty, and secondarily on your region. Sixty-two percent would be worth investigating in office-based primary care, where 45–50% is typical, and unremarkable in a procedural specialty where 60–70% reflects real equipment and staffing intensity. Above 65% in any specialty, check the definition first, then the denominator, then whether collections have fallen — in that order.
Which single line should I look at first?
Staff, every time. It is about 25–30% of collections and roughly half of total overhead — $400,000 of $760,000 in the example. A 10% improvement there is worth $40,000 a year, against $8,000 for the same percentage off supplies. Look at staff-per-provider ratio and scheduling utilisation before you touch vendor contracts.
Does overhead percentage matter when I sell the practice?
Yes, but a buyer will rebuild it rather than accept yours. Valuation normally works from normalised earnings — adding back owner compensation above market, removing discretionary spending, and restating overhead on the buyer's own definitions. A practice whose overhead looks high only because provider compensation was folded into staff costs sees its normalised earnings rise the moment that salary moves back below the line, so fixing the definition is worth real money before any conversation starts.
Sources
- MGMA cost and revenue survey data — the specialty-specific practice-cost reference, available to subscribers and to practices that submit data; figures are not public
- CMS Medicare Physician Fee Schedule — the practice expense RVU, the geographic practice cost indices, and the facility versus non-facility payment differential
- Your specialty society's practice-cost survey, and your own practice's trailing-twelve-month profit and loss statement
- Benchmark bands as implemented in the Investor Sam practice overhead calculator
Benchmarks vary substantially by specialty, region, payer mix and practice model, and the figures here are illustrative. This is not financial, tax or legal advice; review your own statements with your accountant or practice adviser.