Compare · 10 min read
Adding a provider is a cost-structure decision, not a production one. And the advantage runs out, at a point the research can actually name.
Group practices are supposed to run lower overhead than solo practices. On cost per dentist that is true, and the gap is large. On the one measure that charges every dentist for their clinical time, including you, it reverses, in the most recent year the American Dental Association has published that split.
Both of those statements come from the same survey and the same year. That is not a paradox, it is two different measurements, and knowing which one you are looking at is what decides whether adding a provider makes you money. Here is where the scale advantage is real, where it turns, and where it runs out entirely.
General practice owners, 2025 data, per dentist, from the ADA's Survey of Dental Practice. Solo means one dentist in the practice; nonsolo means more than one.
| Solo | Multi-provider | |
| Gross billings per dentist | $1,062,180 | $836,540 |
| Practice expenses per dentist, owner pay excluded | $600,520 | $497,040 |
| Owner's net income | $207,520 | $262,880 |
| Hours per year in the office | 1,729.6 | 1,720.3 |
Read the first two rows together, because separately they each mislead. The solo dentist bills 27% more per dentist and takes home $55,360 less. The multi-provider practice produces less per chair and pays its owner more.
It is not effort. Hours in the office are within 10 a year of each other, which on a 1,720 hour base is noise. It is not contracts either: private insurance accounts for 50.2% of solo billings and 49.2% of multi-provider billings, close enough that the share of billings coming from carriers is not the explanation, though it says nothing about how deep the contracted write-offs run in either model.
What is left is cost structure, and that is the whole story.
Costs per dentist run $103,480 lower in the multi-provider practice. That is a 17% reduction and it is not an accounting artifact. It comes from the part of a dental practice that does not care how many dentists use it.
Add a provider and some costs follow them directly: clinical supplies, lab, assistant chair time, and their own compensation. Other costs do not move at all in the short run. The building does not get bigger. The sterilization area does not double. The practice management software costs what it costs. The operatories already exist, and if they were sitting empty two days a week, the second provider is using capacity you were already paying for.
That fixed base is the dilution, and it is measurable. The only study that estimates a dental practice cost function directly is Chen and Ray at the University of Connecticut, working from an ADA survey of 117 Colorado general practices. They put minimum average cost at 50.6 cents per dollar of gross billings, while marginal cost at the average practice's output is 63.5 cents. Over 90% of the practices in that sample were still on the part of the curve where growing lowers average cost.
Caveat that honestly, because it is the best evidence available and it is thin: 117 practices, one state, the ADA's 2005 to 2006 survey, and a working paper rather than a peer-reviewed journal article. It is directionally the strongest thing in the literature and it should not be quoted to two decimal places at your own practice.
The same study estimates the point at which scale economies are exhausted and average cost starts rising again. Their two methods put it at roughly $1,677,000 and $1,890,000 of gross billings, in 2006 dollars. Carried forward on consumer price inflation that is about $2.7 million to $3.0 million today, and dental costs have outrun general inflation, so treat those as a floor rather than a ceiling.
Set that against the ADA's solo general practice at $1,062,180 of billings. At the $836,540 a multi-provider practice bills per dentist, the distance between where a typical solo practice sits and where the curve turns is, in round numbers, two added providers. Dentist two and dentist three are where the arithmetic works. Past that, on this evidence, you are into the part of the curve where coordination costs more than it saves.
Nobody selling you on growth mentions the second half of that curve. It is also the most useful thing in the research, because it reframes the decision. Adding a provider is not a strategy you keep applying. It is a specific move with a specific range where it pays.
Now the number that cuts the other way, and it is the one to sit with.
The ADA also publishes practice expenses including every shareholder's salary as a percentage of collections. On that measure, incorporated general practice owners in 2025 came in at 94.4% solo and 97.6% multi-provider, on 73 and 53 responses. The multi-provider practice is roughly 3 points worse.
The reason is not mysterious. Once you charge every dentist in the building for their clinical time, including yourself, the associate's compensation shows up as the real cost it is. The multi-provider practice bought its lower fixed cost per dentist partly by purchasing clinical hours the solo owner supplies personally and does not invoice. Some of that $103,480 saving is not a saving. It is a substitution.
One caution on how far to push that. The ADA publishes the solo and multi-provider split of this particular ratio for the current year only, so read the 3 point gap as one reading rather than a trend. What is demonstrably stable is the level. The all-practice version of the same ratio, published for 2002 and then annually from 2010, has sat between 95.2% and 99.6% across all 17 readings with no trend in either direction. A measure that includes owner pay lands near 100% in a good year and a bad one, by construction.
Which is exactly why associate pay and owner pay belong on separate lines. Collapse them and the question you most need answered, whether the associate is actually profitable, becomes unanswerable, and no overhead percentage will tell you.
Two things scale badly, and one of them contradicts the folklore directly.
Staffing grows faster than provider count. In the ADA Health Policy Institute's economic outlook survey covering the end of 2025, 61.8% of group practices with 2 to 9 dentists had added staff since the start of the year, against 36.1% of solo practices. Adding a chair does not add one salary.
The front desk is the first thing to stop being fixed, not the last. In the same survey, 35.8% of group practices had recruited administrative staff in the prior three months, against 16.3% of solo practices. That gap is wider than the gap for hygienists and comparable to the gap for assistants. The usual assumption is that clinical staff scales and admin does not. The data says the opposite: scheduling, insurance verification and collections across multiple providers is where the work compounds, and it compounds before the clinical side does. Practices that run payroll across several providers tend to discover this in the second year, not the first.
Chen and Ray found something in their own sample that fits, and it is worth naming. In their sample, technical efficiency averaged 0.978 while allocative efficiency averaged 0.810. Practices were not working inefficiently. They were buying the wrong mix of dentist hours, hygiene hours, assistant hours and operatories for the prices they faced. That decision gets harder with every provider you add, and it is a plausible reading of why the multi-provider expense ratio comes in higher despite the lower fixed cost per dentist.
Use the same band for both. Total overhead of 55% to 65% of net production, with the owner's own compensation outside it and associate compensation treated as a clinical cost rather than as overhead. One band, applied consistently, beats two bands that let you grade on a curve.
If you want to see where your own lines sit against the full set of bands before you run that comparison, our free Dental Practice Benchmark Scorecard takes about two minutes.
What changes is what you expect to see and what you do about it:
The trap is benchmarking a multi-provider practice as though it were a larger solo practice. It is a structurally different business with a different cost curve, and the fact that both report a percentage does not make the percentages comparable. Look at staffing percentage alongside it in either case, since payroll is the largest category in both models and moves first.
The question is not whether your overhead percentage will fall. On cost per dentist it very likely will, and that is the least interesting thing that happens.
The question is whether the added production covers the added clinical cost plus the added administrative cost, and whether the capacity you are filling was genuinely idle. If your operatories are already full six days a week, a new provider is buying you a building expansion you have not budgeted, and the scale argument does not apply. If two operatories sit dark on Thursdays and Fridays, the fixed base is already paid for and the arithmetic is much friendlier.
Run it as a forecast with real numbers before you run it as a hire. What does the provider produce, what do they cost fully loaded, what does the additional admin support cost, and where does that leave the practice against the band. That is a 30 minute exercise on a clean set of books and an unanswerable question on a messy one.
P.S. Reciprocity Accounting keeps provider compensation on its own lines, so you can read your overhead with the associate in and with the associate out and know which question each answer belongs to. See how we can help your practice.
On cost per dentist, yes, and substantially. ADA data for 2025 shows practice expenses per dentist of $497,040 in multi-provider general practices against $600,520 in solo practices, with owner compensation excluded from both. On expenses including every dentist's pay as a share of collections, it reverses: 97.6% multi-provider against 94.4% solo. Both figures are correct and they answer different questions.
Partly real and partly a denominator effect. Multi-provider practices count associates in the denominator, and in our experience associates are often newer, sometimes part time, and still building a patient base. Hours worked are nearly identical between the two models, so it is not an effort difference. It is not a one year artifact either. The ADA's median gross billings table carries the solo and nonsolo split back to 1990, and solo runs higher in every year of it.
The only direct estimate available puts it at roughly $1.68 million to $1.89 million of gross billings in 2006 dollars, or about $2.7 million to $3.0 million today, above which average cost rises. Measured against a typical solo general practice at about $1.06 million, that is roughly two added providers. Treat it as directional: the estimate comes from 117 practices in one state from the ADA's 2005 to 2006 survey.
No. Associate compensation is a clinical cost and belongs with doctor pay, alongside the owner's. Putting it in overhead inflates the ratio and hides whether the associate is profitable, which is a separate question that an overhead percentage cannot answer either way.
The same 55% to 65% of net production, with owner compensation excluded and associate compensation in clinical cost. The temptation is to hand a group practice a looser band because its ratio comes in higher, and that is exactly the wrong move. The higher ratio is the finding. Widening the band to accommodate it deletes the only signal the comparison was going to give you. One band, both models, and let the difference show.