My first question was for their year-to-date numbers. I’m never surprised anymore when it comes to numbers. It was almost expected that I would find the collections were only up .9%.

His preplanned explanation came next. The waitlist showed demand, and the flat collections were a posting problem. He said somebody in the front office must be falling behind on payments. Once again, the explanation he gave was expected. I’ve heard it before. Chalk it up to 30 years’ experience or one of my many personality defects. I quit trying to figure out the difference years ago.

The schedule is real. The story about it is wrong.

A full book does measure something. It’s just not what he thought.

The fee schedule, like almost all fee schedules in general dentistry, is mostly dictated by the plans they participate in. Those plans have been paying less in real terms every year for five years. The ADA’s Health Policy Institute puts the reimbursement rate index up 19% since January 2021. General inflation over that same stretch ran 27%. In 2026 alone, January through June, reimbursement moved 1.1% while inflation was 1.8%.

Look at it this way: when a price falls in real time for five years, the chairs are filling with the work that price underprices. The patients are doing the sorting, which means they are keeping the cleaning and the exam. The emergency gets seen because it’s an emergency and it has to. But the restorative work is getting pushed to next quarter, and the implant is getting pushed out a year until the household feels better about the money. The schedule is getting filled with visits, and each visit is carrying less. (Hopefully that came out right. My voice-to-text method receives much less refinement since someone flagged me for AI slop on LinkedIn. That’s another essay.)

Now you and he know what the schedule was actually measuring. A queue forms at a price that’s too low for the mix. He was reading rationing as growth.

One practice, two years

Treat the practice below as a composite, built from the shape I keep seeing. The numbers are rounded so the math can be checked on a napkin. 5 operatories, 8 hours per day, 4 ½ days per week, 48 weeks per year.

Operatory hours, both years = 5 x 8 x 4.5 x 48 = 8640

Last year:

Hygiene and exam visits = 6,000 at $190 realized = $1,140,000

Restorative visits = 3,000 at $520 realized = $1,560,000

Collections = $2,700,000 Visits = 9,000

This year:

Hygiene and exam visits = 6,800 at $192 realized = $1,305,600

Restorative visits = 2,700 at $525 realized = $1,417,500

Collections = $2,723,100 Visits = 9,500

Visits up 5.6%. Collections up 0.9%. 500 more people came through the door, the wait list got longer, and the money barely moved. That’s what was actually being presented.

Now take the collections change apart. Only three things can move it, and one of them is doing most of the work here.

Volume effect = 500 more visits × $300 last-year average = +$150,000

Mix effect = (6,800 × $190 + 2,700 × $520) − (9,500 × $300) = −$154,000

Price effect = $2,723,100 − $2,696,000 = +$27,100 Net = +$23,100

The .9% he trusted is a $150,000 gain and a $154,000 loss standing on top of each other. The practice grew and shrank in the same year, and the P&L netted the two into a rounding error.

Then run contribution, with the hygienist raise in it, because he gave one this year and so did everybody else.

Hygiene contribution per hour, last year = $190 − $60 − $10 = $120

Hygiene contribution per hour, this year = $192 − $62 − $10 = $120

Restorative contribution per hour, last year = $520 − $30 − $110 = $380

Restorative contribution per hour, this year = $525 − $31 − $112 = $382

Total contribution, last year = 6,000 × $120 + 3,000 × $380 = $1,860,000

Total contribution, this year = 6,800 × $120 + 2,700 × $382 = $1,847,400

Contribution per operatory hour, last year = $1,860,000 / 8,640 = $215

Contribution per operatory hour, this year = $1,847,400 / 8,640 = $214

Same practice. 500 more visits. $1 less per hour the chairs were occupied. See how he’s working harder for the same money? And the schedule that told him things were going well is the reason.

Why the chairs filled with the wrong work

Two forces did this, and a third explains why he couldn’t see it.

The first is the price. A fee schedule that lags cost for five years stops sorting. Above a clearing price, a practice chooses its work. Below it, the patients choose the cheap visit. HPI’s own report has the second most common reason pessimistic dentists give for their outlook as patients unwilling or unable to prioritize care. That’s what’s called mix moving.

The second is that the fixed hour got more expensive while the fee for filling it didn’t move. The hygienist raise is in the math above. It took $2 per visit off the hygiene contribution, the fee gave back $2, and the department came out flat while looking busier than ever.

The third is the reporting, and this is the one that keeps the other two from being seen. Practice management software reports production and collections as totals because that’s what comp gets paid on. Nobody in the practice gets paid to look at contribution by procedure class per occupied chair hour, so nobody does. The number that would have flagged the mix shift in March is the number none of his reports produce. (If you follow my work, you’ll recognize the metrics. Contribution per owner hour after collections per matter, contribution by procedure class per occupied chair hour.)

The number the software won’t print

Contribution per occupied chair hour, by procedure class, monthly. I’ve defined it here so you can run it yourself:

Contribution per occupied chair hour = (collections − provider and hygienist labor − assistant labor − supplies and lab, by procedure class) / occupied chair hours

Occupied, not scheduled. Hygiene against restorative against elective. Run it by class, and the schedule template turns into a P&L, because every hour you allocate to one class is an hour you took from another at a known difference in contribution. (More Sheets language.)

Pair it with the decomposition above, and run monthly. Volume, mix, price. The mix line is the early warning. It goes negative months before collections notice.

The sixth operatory was the wrong question

He appeared in my DM’s asking the wrong thing. The question I answered was what the fifth operatory produces per hour today, by class, and what a scheduling rule that moves hours between classes would do to contribution before it goes flat.

Let’s look at the levers here. At a fee schedule he doesn’t control, the lever he does control is the template. Which hours go to which class of work. Behind it sits case acceptance on plans above a dollar threshold, and further back, the question of which plans he keeps at all. Each of those is a rule, and a rule holds (think: a governed decision system) when a front desk, under pressure, would otherwise book whatever fits. RIDA (the economic discipline I created) runs the decomposition first and writes the rule second.

There’s a transaction version of this, and it is worth mentioning here. The CIM, if he ever writes one, will say fully booked with a waitlist, because it’s the truth and it sounds like demand. The buyer’s operating partner will run contribution per occupied chair hour on the second day of diligence and then price the growth story.

A seller who ran it first has an answer.

This essay describes general transaction structure. It is general in nature and does not constitute legal advice, a valuation, or investment advice. Figures drawn from published sources are attributed in the text; the two-year figures are illustrative composites and describe no identifiable practice.

Sources. ADA Health Policy Institute, The State of the U.S. Dental Economy, Q2 2026 (July 2026): reimbursement rate index up 19 percent since January 2021 against 27 percent general inflation; January through June 2026 reimbursement up 1.1 percent against 1.8 percent inflation; share of dentists reporting not busy enough at 26 percent, down from 33 percent in Q4 2025; average new-patient wait 13.9 days; 589 responses, 552 in private practice, owner-skewed. U.S. Bureau of Economic Analysis: real consumer spending on dental services up 1 percent over the twelve months to May 2026. The two-year practice above is a composite; figures are rounded and illustrative.

The first few times I took that call, I went looking for a leak. I never found one. The money was all there and all accounted for. It was sitting in her clients’ bank accounts instead of hers.

The P&L measures when work was earned. Cash arrives later. In a growing firm the gap between those two dates widens every year, because every new matter opens with labor and closes with a check months down the road. That’s why the squeeze shows up in the good years. The better the year, the more the firm has earned that it hasn’t collected.

The industry has a name for the delay: lockup, the days between doing the work and banking the money. Clio’s Legal Trends Report puts the median firm at about 93 days of it. Three months of finished work, earning nothing.

What the delay was costing her

Her firm collected $2.6M that year. Start with the daily rate.

$2,600,000 ÷ 365 = $7,123 collected per day

Her lockup ran near the median, about 95 days.

95 days × $7,123 = $676,685

That’s the balance of finished work sitting outside her bank account on any given day, all year, as a standing feature of how the firm operates.

She’d earned every dollar of the record. She’d collected about ten months of it.

I’ve worked with firms at the same collections that get the cycle down to 25 days. Bills go out on the first of the month and follow-up starts at day thirty.

25 days × $7,123 = $178,075

$676,685 − $178,075 = $498,610

Half a million dollars, and the difference between the two firms is billing habits. Clio’s quartile data says the spread is real in the wild: the best firms carry under 20 days of unbilled work while the bottom quartile carries more than 78.

Now price the gap, because her bank already had.

$498,610 × 9% = $44,875

Nine percent is a composite line-of-credit rate for the demonstration; her figures are composites drawn from engagement work and rounded. She was paying her bank roughly $45K a year to finance credit she was extending to her clients at no charge. The interest showed up on her P&L as a finance cost. Nothing on the statement connected it to the prebills sitting in partner review.

There’s a second cost underneath the interest. The loan doesn’t repay at par. Clio’s numbers again: about 14% of billable work at solo and small firms never gets invoiced, and roughly a tenth of what’s invoiced never gets paid. Old WIP is where most of that dies. The longer the work sits, the more of the principal never comes back.

Why her reports couldn’t see it

Start with the accrual layer. The P&L answers one question: did the firm earn money. It answers correctly. Nobody built a report to track how much the firm has lent out, because nobody decided to lend anything.

Which is the second layer. The delay accumulates one postponed billing run at a time. WIP waits on prebill review, and statements go out whenever someone gets to them. No target was ever set, so no report can show a miss.

The third layer is the one that matters. A firm carrying 95 days of lockup is in the lending business. Every engagement letter extends unsecured credit at 0% as a default term. No borrower was underwritten and no limit was set. A bank that ran its book this way would have regulators in the lobby. A law firm that runs its book this way has a record year and a February cash problem.

The number to pull this afternoon

Your practice management system already holds both inputs.

Lockup days = days from work performed to bill sent + days from bill sent to cash received

Loan balance = lockup days × (annual collections ÷ 365)

Annual carry = loan balance × your line-of-credit rate

Run it on your own book. Your reporting produces a profit figure every month. It has never produced this number, and this is the one that explains February.

What a priced loan looks like

There’s nothing exotic about the fix, which is the part that surprises people. The firms at 25 days aren’t running better software than the firms at 95. They wrote the loan terms down. Billing goes out on a calendar instead of a mood. A lockup target sits in the monthly partner report next to origination, so the delay finally has a number a meeting can look at. Some firms go further and price the credit directly, with deposits held against work in process, or evergreen retainers that refill as they draw down. Slow payers get shorter terms, the same way they would at any other lender.

None of that is collections pressure. The clients were always going to pay. The open question was when, and a firm that puts the answer in writing usually finds most clients had no opinion about it at all. Nobody’s client asked for 95-day terms. The firm offered them without noticing.

The partner asked me why cash was tight in a record year. The question that governs is different: what does the firm charge for the credit it extends, and on what terms. At the median firm the answer is nothing, on terms the clients set.

A firm that never set a credit policy still has one. The clients wrote it.

The client figures are composites drawn from engagement work, labeled as such. Market figures are from the Clio Legal Trends Report, paraphrased and attributed: median lockup of roughly 93 days; top-quartile firms under 20 days unbilled against bottom-quartile firms past 78; about 14% of billable work never invoiced and roughly a tenth of invoiced work unpaid.

Sources. Clio Legal Trends Report (lockup, realization, and collection benchmarks).

If the number you pulled is bigger than you thought. The Growth Intelligence Scorecard reads your firm’s revenue structure from the numbers you already carry. About four minutes, in your browser, no meeting. growthprolegal.com/scorecard

In February, the owner of a three-doctor small-animal practice collecting $2.8M approved the annual fee increase. 6% across the schedule, loaded into the practice software March 1. She had the regional benchmarks, the appointment book was full, and nobody on the team pushed back. By June, revenue per invoice was running below the prior year. The fee schedule went up. The prices came down. Nobody edited the schedule in the software, and nobody had to. The one field that grew was the discount field. The practice is a composite drawn from engagement work; the arithmetic below is the arithmetic that surfaced when we decomposed it.

The number the owner trusted was the fee schedule. It records what the practice intends to charge, it gets reviewed once a year against the benchmarks, and it feels like the pricing policy because it has the fees printed in it. The document that actually sets the practice’s prices is the compensation plan. The schedule states an intention. The comp plan decides, invoice by invoice, what the practice accepts, because the comp plan is the document that tells each doctor what a discount costs her personally. In most owner-run practices those two documents were written in different years, for different reasons, and reviewed by different people. When they disagree, the comp plan wins. It is the one attached to money.

One estimate, two answers

Take a $2,400 dental treatment plan at schedule fees, extractions included. The associate is paid 20% of her production, the bottom of the 20 to 22% band that practice appraisers and the ProSal literature both describe as standard for companion-animal general practice. The client winces at the estimate, and the associate reads the room: hold the number and the plan gets approved six times in ten; take 10% off and it gets approved seven and a half times in ten. Composite figures, rounded for the demonstration, and the structure survives any reasonable substitution.

The associate’s decision. At full fee her production credit pays $480. At the discounted $2,160 it pays $432. Holding: 0.60 x $480, an expected $288. Conceding: 0.75 x $432, an expected $324. The discount is worth $36 to her. She grants it, and on her math she should.

The practice’s decision. Delivering the dental costs $1,300 in technician time, anesthesia, drugs, and allocated overhead. At full fee the practice clears $2,400 less $480 in doctor comp less $1,300 in delivery: $620. Discounted, it clears $2,160 less $432 less $1,300: $428. Holding: 0.60 x $620, an expected $372. Conceding: 0.75 x $428, an expected $321. The discount costs the practice $51 in expectation. On the practice’s math, the number holds.

Same patient. Same client. Same probabilities. Opposite correct answers. Neither party ran the math wrong, and nobody misbehaved. The comp plan handed one pricing decision to two parties and turned it into two different problems, and the person standing in the exam room is solving the one that pays her. A single case is small money. A doctor makes this call several times a day, every doctor in the building makes it, and the exam room is exactly where the fee schedule has no one to defend it. That is how a schedule rises 6% while revenue per invoice falls.

Why the plan prices this way

The plan splits every discount unevenly. A 20% production share means the doctor absorbs 20 cents of each discounted dollar and the practice absorbs 80. That $240 concession cost her $48 and the practice $192. This is incidence, the oldest question in price theory: who actually bears a cost, as opposed to who nominally grants it. Whoever bears the smaller share of a price cut will grant more of them. That follows from the split. The instinct is to see a soft-hearted associate who can’t hold a number in front of a crying client. The split predicts the behavior regardless of who is in the room, which is why replacing the associate changes nothing. Mark Opperman, whose ProSal model is the ancestor of most of these plans, has estimated for years that a discounted dollar takes roughly four dollars of new revenue to replace. The warning and the plan have coexisted for decades. The incidence arithmetic above is the reason the warning keeps losing.

Production comp is blind to margin. The plan pays the same 20 cents whether the dollar cost the practice 55 cents to deliver or 85. A doctor paid on gross production will rationally pursue gross, and the practice’s profit lives entirely in the spread the plan can’t see. The services with the thinnest margins are often the ones where a client balks and a discount closes the gap, so the blindness compounds: the plan steers its holders toward conceding exactly where a concession does the most damage.

And a production share prefers volume at any workable fee. More approved plans always pay the doctor more. The practice runs on finite doctor hours and finite surgery slots, and the discounted dental consumes the same anesthesia block as the full-fee one it displaced. Capacity is the constraint the plan never met. AVMA benchmarking has the average practice losing roughly 95 active clients a year since 2019, and a shrinking client file is exactly the backdrop that makes a walk-out feel expensive in the exam room and a concession feel cheap. The plan and the fear point the same direction.

The number your reporting can’t produce

Call it the discount incidence report. For every doctor with discretion at the invoice: twelve months of her discounts, waived line items, and comfort adjustments, each split by the plan into her share and the practice’s share. Your software already totals the discount field, and that total is the practice-wide leak. It doesn’t show whose hand granted it, and it doesn’t show what fraction of each conceded dollar that doctor personally paid. Those two columns are the ones that explain the behavior. At the composite practice, roughly 70% of the year’s discounting traced to two of the three doctors, and their personal share of the dollars they conceded was 20 cents on the dollar. They were spending practice money at a four-to-one exchange rate, with authorization, per the plan. The report takes an afternoon to build from data the software already holds. Most owners have never sorted the leak by the hand that opened it.

The governing question

The owner arrived asking whether the fees were right. That was the wrong question. The fees were fine; the benchmarks said so, and clients kept approving full-fee plans six times in ten. The governing question is who holds pricing authority under the comp plan, and what each concession costs the person holding it. Get that split wrong and no fee schedule survives contact with the exam room. This is why RIDA, the proprietary economic discipline B.L. Sheets & Co. runs on, sequences decision architecture ahead of any fee work: pricing guardrails only hold when the comp plan stops paying people to breach them.

The fix is rarely dramatic. Sometimes it is a production credit computed on contribution instead of gross. Sometimes it is a discount threshold above which the concession comes out of the doctor’s credit at the practice’s ratio instead of her own. In every version, the move is the same: make the person who sets the price bear the price’s consequences in the proportion the practice does. Alignment does the enforcement, and the fee schedule starts meaning what it says.

A practice prices the way it pays.

Subscribe now

Sources

Mark Opperman, CVPM, in dvm360, on the ProSal compensation model, standard associate production percentages, and the revenue cost of discounting. Byron Farquer and David McCormick, What It’s Worth: Veterinary Practice Value, on small-animal associate compensation at 20 to 22% of production. AVMA Veterinary Economics Division benchmarking, 2024 data, on active-client trends. The practice in the demonstration is a composite drawn from engagement work, and its figures are labeled as such in the text.

The Decision Layer publishes on Thursdays. Each essay takes a number the operator trusts and shows what it was actually measuring.

Both $650,000 figures are real. Both practices collected the money, paid the bills, banked the difference. What neither number discloses is the price it used for the most expensive input in the building, which is the owner’s own clinical work. A dental P&L prices every input at market except that one. Staff wages are market. Rent gets restated to market in about a minute, buyers routinely reset related-party rent to the 5 to 7% of collections that market leases run, per McLerran & Associates’ 2026 valuation guidance, and nobody argues about it for long. Supplies are market. The owner’s chair time is priced at whatever the owner’s accountant decided was optimal that year. A buyer who bids off the presented EBITDA is bidding on the seller’s tax plan.

None of this is obscure. FOCUS Investment Banking’s 2026 dental valuation work lists owner compensation adjustment to fair-market provider rates as part of the standardized normalization every general dentistry deal runs through, and McLerran calls that adjustment the largest single add-back in dental valuations, bigger than personal expenses, which typically run $30,000 to $150,000 a year, bigger than related-party rent. The correction is standard. What varies is when a buyer runs it. Before the indication, or after the LOI has already anchored the price.

Same bid, two deals

Both owners personally produce $1.4 million of the $2.4 million in collections. A general dentist producing at that level costs 30 to 35% of collections to employ; the ADA’s associate compensation guidance and the recruiting market both put the range there, and it’s held for years. Call it 32%, fully loaded. $448,000 is what the acquirer will pay a contracted dentist to keep that production after close, and nothing in either seller’s P&L can change it.

Practice A’s owner pays herself like an employee: $500,000 in W-2 comp, above the line, eaten by the P&L before EBITDA is computed. Add it back, deduct the $448,000 replacement cost, and EBITDA at replacement cost comes to $702,000. Her presentation understated the practice. The $3.9 million bid is 5.6x real earnings, inside the 5 to 8x add-on band FOCUS publishes for general dentistry.

Practice B’s owner pays himself like a tax plan. Salary set at $140,000, defensible as reasonable compensation, and the rest of his take flowing out as distributions below the line, where EBITDA never sees it. So his P&L carried $140,000 of cost for $1.4 million of clinical production, year after year, and the margin everyone in his life has praised, his banker, his study club, the brokers who cold-call him, was partly his own unpaid wages. Add back the $140,000, deduct the same $448,000, and EBITDA at replacement cost is $342,000. The same $3.9 million bid is now 11.4x real earnings. That’s above the 9 to 11x FOCUS publishes for institutional-grade platforms, paid for a single-location add-on whose owner still produces 58% of collections.

Same bid, two deals. One is a fair add-on at 5.6x. The other overpays by roughly $1.8 million against the group’s own 6x applied to the real number, and both answers were sitting in the payroll register and the production report, two documents that arrive with the CIM.

Where the margin came from

Start with why the number exists. The S corporation election that dominates practice ownership rewards a low W-2 salary, since payroll tax applies to salary and spares distributions, and every year the election runs it writes a labor subsidy straight into the margin. Nobody is hiding anything. The number was built by the seller’s accountant for a specific audience, the IRS, and it did its job. It’s now being read by a second audience it was never built for, and the second audience has a spreadsheet.

That spreadsheet carries one cell for the owner’s clinical labor: what the production costs to replace, because the day after close it has to be bought from a dentist with a contract and a market rate. McLerran puts the sensitivity plainly. A $100,000 difference in adjusted EBITDA moves enterprise value by $700,000 to $900,000 at the 7 to 9x range where competitive processes land. Practice B’s gap was $360,000.

There’s a second exposure riding on the first. The owner whose underpriced labor inflated the margin is usually also the owner whose production concentration threatens it, and that one is priced separately. Sofer Advisors’ provider risk analysis, carried in FOCUS’s 2026 work, puts practices with owner production of 90% or more at valuation reductions of 10 to 20%, and provider risk moved from a consideration to a primary decision driver this cycle. Miss the compensation subsidy and you’ve probably also mispriced the person the subsidy walks out with.

Then there’s sequence, which decides who pays for the error. Run before the indication, the normalization costs an afternoon and reprices the bid. Run after the LOI, it’s a retrade. Exclusivity burning, the seller anchored to the headline number, the diligence spend committed. The ADA Health Policy Institute has corporate dental affiliation at 16.1% of U.S. dentists in 2024, up from 7.2% in 2015, which is a market full of institutional buyers running this same playbook, and the one who normalizes late is the one funding the seller’s anchor. RIDA, the proprietary economic discipline B.L. Sheets & Co. runs on, treats this as Stage 1 work, cost separation: the owner’s labor priced as labor before the margin gets read as margin.

The number the CIM will never hand you

EBITDA at replacement cost. It’s computable before the bid, from documents already in the data room. Take presented EBITDA. Add back every dollar of owner clinical compensation expensed above the line, salary, payroll taxes, benefits. Deduct the market cost of replacing the owner’s clinical collections, 30 to 35% of them for a general dentist, fully loaded, more for specialists. Per owner, if there’s more than one. The result has no fixed relationship to the presented figure. It sat above it at Practice A and 47% below it at Practice B, and the direction was knowable from the payroll register before anyone drafted an LOI.

Run it and the question changes. A buyer arrives at a deal book asking whether the margin is real. One layer under that sits the question that decides the deal: whose price list computed it. A margin computed on the owner’s tax elections answers a question the IRS asked. A margin computed at replacement cost answers the one the acquisition is asking. Until the second computation exists, the bid is priced on the first, and only the seller’s accountant knows by how much. An indication priced on the second number costs nothing extra to produce. Sellers with real earnings clear it untouched, and sellers whose margin was a payroll election get repriced on paper instead of in exclusivity.

What the bid was buying

The practice was profitable at the owner’s price for the owner’s labor. The buyer inherits the market’s price on day one. A bid built before that repricing buys the subsidy at a multiple, and the subsidy resigns at close.

Keep Reading

The DSO Offer That Looks Bigger Than It Is

Owner Dependency Is the Discount

Subscribe now

The Decision Layer publishes weekly on the revenue, capital, and incentive decisions that determine whether a firm compounds or quietly comes apart. This essay is general commentary. It is not legal, valuation, tax, or investment advice, and no figure here describes any identifiable practice. The two-case figures are composites and are labeled as such.

Sources: FOCUS Investment Banking, Dental Practice EBITDA Multiples 2026 report and 2026 dental valuation guidance (multiple ranges, standardized normalization, provider concentration discounts); McLerran & Associates 2026 DSO valuation guides (owner compensation as the largest single add-back, personal expense and related-party rent adjustment ranges, EBITDA-to-enterprise-value sensitivity); Sofer Advisors provider risk analysis, as carried in FOCUS’s 2026 work (owner production concentration and valuation reductions); ADA Health Policy Institute (corporate dental affiliation, 2015 to 2024); ADA associate compensation guidance (associate market at 30 to 35% of collections). Two-case figures are composites.

The practice’s best line and its biggest risk are the same line, and the line is the owner.

A buyer prices what transfers

The cause is structural. A practice’s earnings blend two components the P&L never separates. One part is produced by the system: the associates, the technicians, the protocols, the recurring wellness revenue, the client base attached to the building. The other part is produced by the owner personally, by her hands and her relationships. The blended EBITDA reports the two as one number. A buyer separates them, because only one of the two is for sale. The owner’s production sits in the trailing twelve months, and in three years it will be wherever the owner is. A buyer prices the future cash flow that survives the transfer, and cash flow produced by the seller’s own hands survives only if it can be replaced.

The market prices this separation openly. Published 2026 veterinary transaction guides put associate-driven general practices at 9 to 13 times adjusted EBITDA and owner-dependent practices at 4 to 7, and the sector literature treats owner clinical share as the single largest multiple driver in veterinary M&A, ahead of size, geography, and payer dynamics. Practices with owner production above 70 percent are observed at the bottom of the band for their size. Valuation practitioners put the provider concentration discount at 20 to 30 percent on its own. The discount arrives before negotiation begins. It is built into the band the buyer selects on day one.

Same EBITDA, different money

Two three-doctor general practices in comparable markets, each collecting $2.9 million, each showing $520,000 in adjusted EBITDA. On the financials, the same practice twice.

In the first, the owner produces 30 percent of collections. Two seasoned associates carry the rest on signed agreements, the surgical schedule is distributed, wellness plan revenue recurs monthly, and the busiest clients have relationships with the practice, because the practice made sure of it. A consolidator’s indication arrives at 8 times: $4.16 million, with a standard one-year clinical transition for the seller.

In the second, the owner produces 65 percent. The associates handle wellness and routine medicine while she performs all the surgery and dentistry, and the referring clinics in the area send cases to her, by name. The indication arrives at 5.5 times: $2.86 million, conditioned on a 24-month post-close employment commitment, because the buyer’s model treats her exit as the loss of $1.9 million in annual production. Replacing that production means recruiting roughly two full-time doctors in a market where associate compensation averages about $130,000 and runs to $165,000 before benefits and signing incentives, and where, per AVMA and industry workforce reporting, associate seats are the hardest roles in the profession to fill and stay open the longest. The buyer prices the cost of that replacement and the odds of achieving it, and both go into the multiple.

Same collections, same EBITDA, $1.3 million apart before a single term is negotiated. The gap is the discount, and it was set by a number neither practice reports.

Three forces, each deeper than the last

The first force is concentration itself. The revenue is real, but ownership of it is split. Revenue produced by the system belongs to the practice and transfers with the keys. Revenue produced by the owner belongs, in the economic sense that matters to a buyer, to the owner, and the sale of the practice does not include her. The higher her share, the smaller the business actually on the table, whatever the top line says.

The second force is replacement scarcity, and it is what converts a staffing question into a valuation event. If replacement doctors were abundant, owner production would be a line-item cost, the market wage times the hours. They are not abundant. The profession’s workforce gap is structural, demand for clinical capacity is growing faster than the supply of doctors, and the major consolidators are competing for the same associates every independent practice is trying to hire. So the buyer prices two things: the full market cost of the replacement, and the risk that the search takes eighteen months or fails. Scarcity moves the discount from the cost of a salary to the price of an uncertainty, and the buyer holds the pen on pricing it.

The third force is attachment, and it is the one no hire can fix. Veterinary medicine is relationship medicine. A meaningful share of an owner-dependent practice’s goodwill is personal: clients bonded to the doctor, referral flows aimed at her name. Personal goodwill transfers only by handoff, slowly, doctor to doctor, visit by visit, which is exactly why buyers of owner-dependent practices require the 12 to 24 month retention periods that sector advisors describe as the standard structure for narrowing the discount. The owner’s excellence built value the practice cannot own. That is the deepest version of the problem, and it is also the reason the discount feels so unjust to the owner: the thing being discounted is the thing she is best at.

The number the practice cannot report

The measurement is transferable EBITDA: the earnings that survive the owner’s exit. Compute it directly. Take the owner’s clinical production and replace it at the full market cost of the doctors required to produce it, compensation, benefits, recruiting, and ramp time included. Then haircut the revenue attached personally to the owner, the ask-for-her clients and the name-directed referrals, by an honest transition attrition rate. What remains is the practice a buyer is actually pricing. The spread between book EBITDA and transferable EBITDA is the dependency discount, and it is computable years before any buyer exists. No practice management report will ever produce it, because every report the practice runs measures production, and production is precisely the number that conflates the system with the owner.

Run the spread and the owner’s question changes. What the practice is worth turns out to be the wrong question, because there are two practices in the building, and only one of them is for sale. The governing question is how much of the practice leaves when the owner does, and that question rewards early answers. The dependency deepens by default: the owner keeps the surgeries because she is fastest, keeps the top clients because they ask, and every year of that concentrates the practice further into her hands. Closing the spread takes years, not months. A doctor has to be recruited in a scarce market, seasoned, and handed relationships visit by visit, and a buyer wants to see the new distribution hold in the trailing numbers before pricing it. The owner who starts three years out sells the practice. The owner who starts at the sale sells herself, on a two-year employment agreement, at a discount.

What she built and what she can sell

The owner built the practice by being the best doctor in it, and the buyer discounts the practice for the same reason. Both are correct. The production that built the practice is the one asset inside it the owner cannot sell, and the work of a transition is moving as much of it as possible into hands that stay.

This essay describes general transaction structure. It is general in nature and does not constitute legal advice, a valuation, or investment advice. Figures drawn from published sources are attributed in the text; case figures are illustrative composites and describe no identifiable practice.

Sources. Ackerman Group, quarterly veterinary market updates, 2025 through Q1 2026. Provident Healthcare Partners, veterinary sector commentary. AVMA workforce and practice data, 2025 through 2026. VHMA compensation and staffing benchmarks. SovDoc 2025 and published 2026 veterinary transaction guides (DVMElite, Transitions Elite, RightFit Capital) on multiple ranges and owner-dependency effects. ZipRecruiter national associate veterinarian compensation data, May 2026. Industry workforce reporting on the structural DVM shortage, 2025 through 2026.

A general dentist in the Southeast is holding a letter of intent. Page one prices the practice at 7.2 times adjusted EBITDA, a little over eight million dollars. The number fits the market. FOCUS Investment Banking’s 2026 benchmarks hold general dentistry at 9 to 11 times for platform acquisitions and 5 to 8 times for add-ons, and Provident Healthcare Partners’ Q1 2026 sector update shows platform-tier targets above ten million in adjusted EBITDA trading at 10 to 15 times. The letter is real, the buyer is funded, and the owner reads page one and hears eight million dollars.

Pages two through twelve describe a different transaction. They describe how much of the number arrives as cash, how much converts into stock in the buyer, how much is conditional on performance after the sale, and what the owner is agreeing to keep doing, at what compensation, for how long. Page one and pages two through twelve are the same offer.

The multiple prices an assumption

Adjusted EBITDA is a constructed number, and the largest construction inside it is normalized owner compensation. The buyer strips out whatever the owner actually paid themselves and substitutes the market cost of a provider who could produce the same dentistry. If an owner collected 1.8 million dollars in personal production and took 650,000 dollars in compensation, and the market cost of an associate producing 1.8 million is 450,000, the normalization adds 200,000 to EBITDA. The DSO is indifferent to what the owner paid themselves. It prices what the production costs to replace.

The construction cuts both ways, and in this market it mostly cuts down. Per Sofer Advisors’ 2026 analysis, practices where the owner performs 90 percent or more of production see valuation reductions of roughly 10 to 20 percent, and provider risk moved from a consideration to a primary decision driver this year. It was among the top reasons DSOs walked away from signed processes in 2025. Owner compensation is also the largest add-back category in dental quality of earnings work, ahead of related-party rent and personal-use expenses. So the multiple on page one is applied to a number the diligence team hasn’t tested yet, and the multiple survives only as well as the number does.

Same letter, different money

Two owners in the same region each receive a letter at 7.0 times on 1.0 million dollars of adjusted EBITDA. Both hear seven million dollars.

The first owner produces 55 percent of collections. Two associates carry the rest, and her compensation was documented against market rate before the process began. Quality of earnings confirms the 1.0 million. The structure runs 80 percent cash at close and 20 percent rollover equity, no earn-out, with a working capital peg set off a trailing twelve months her advisors verified. The wire at close is 5.6 million dollars, plus a documented 1.4 million minority position in the platform.

The second owner produces 85 percent of collections. Diligence reprices normalized compensation against the true replacement cost of that production, and tested EBITDA lands at 850,000. The same 7.0 multiple now produces a 5.95 million headline. The structure runs 60 percent cash, 30 percent rollover, and a 595,000 dollar earn-out keyed to revenue over 24 months under DSO management. Per SRS Acquiom’s claims research, earn-outs outside life sciences pay about 21 cents on the dollar, which prices his contingent piece near 125,000 at expected value. The working capital true-up then deducts at close against a peg the buyer set. The money that reaches him at close runs a little under 3.5 million dollars.

Both letters carried the same first page. The first held up because the number underneath it had already been built to survive testing. The second was an opening position.

Four places the number moves

Rollover equity converts a sale into a second bet. The rolled portion is an illiquid minority position in a leveraged platform, and its value depends on a future recapitalization at a higher multiple. The market believes in that event, with 78 percent of buy-side respondents in 2026 surveys anticipating recaps within 12 to 36 months, but the seller carries the outcome either way. The distressed dental platforms of 2024 and 2025 took their sellers’ rollover positions down with them.

Earn-outs make part of the price conditional on performance the seller no longer controls. SRS Acquiom’s data puts the median earn-out at 24 months and roughly 31 percent of closing payments when present, with 62 percent keyed to revenue. After close, the schedule, the fee mix, and the staffing that drive that revenue belong to the DSO. Some dental structures now add clawbacks that can require repayment of amounts already received.

Working capital true-ups move the price at close. Per SRS Acquiom’s 2026 study of more than 1,500 private-target deals, working capital adjustments now appear in over 90 percent of transactions, and recent studies report a pro-buyer negative adjustment in 55 percent of deals against a pro-seller positive one in only 35 percent. The peg is a negotiated number, and the party that models it usually wins it.

Post-close terms change what the sale is. The selling doctor typically signs a three to five year employment agreement, and published composites show post-acquisition compensation falling 15 to 25 percent, in one worked example from 420,000 dollars to 310,000. A purchase price that requires three years of below-market employment contains a wage concession the first page never states.

The measurement

The number that governs the decision is net realized proceeds: cash at close, plus contingent consideration priced at expected value, plus the rollover held at an honest illiquidity-discounted mark. Legal-side summaries of DSO transactions put the gap between headline valuation and actual proceeds at 30 to 50 percent once structure is priced. No letter of intent computes this number, because the letter’s job is to state the maxima. Computing it is the seller’s job, and the useful moment to do it is before exclusivity, while there is still more than one buyer in the room.

There is a point in every process where normalized compensation and the add-back schedule stop being a formality and start being the deal. Where that point sits for a specific practice is diligence work. That it exists is structural, and the owners who fare best locate it before the buyer does.

The wrong question

The question owners ask is whether the multiple is good. The multiple is the one number in the letter the seller cannot spend. The governing question is how much of the headline survives the four adjustments, and the answer is mostly decided before the letter arrives, in how the compensation normalization is documented, how the working capital history reads, and how much of the production walks out the door with the owner. An owner who signs exclusivity on the strength of page one spends the next ninety days defending a number the other side constructed.

The first page of a DSO letter states the price of your attention. The other eleven state the price.

The Decision Layer publishes weekly. New essays arrive by email when you subscribe.

This essay describes general transaction structure. It is not legal advice, not a valuation, and not investment advice. Figures drawn from published sources are cited below; illustrative case figures are composites and describe no identifiable practice or firm.

Sources

FOCUS Investment Banking, 2026 dental industry valuation benchmarks. Provident Healthcare Partners, Dental Services Sector Update, Q1 2026. Skytale Group, 2026 Dental M&A Report. Sofer Advisors, 2026 analysis of provider risk and owner-production concentration in dental transactions. SRS Acquiom, M&A Claims Insights Report (earn-out payment rates and structure data) and 2026 M&A Deal Terms Study (working capital adjustment prevalence and direction). Becker’s Dental Review and Dental Economics, reporting on DSO deal structure trends, post-close compensation, and distressed platform outcomes, 2024 through 2026.

A managing partner pulls the Q1 proposal report. Forty-seven proposals were sent; thirty-one won. The discounted proposals closed at 64 percent. The full-price proposals closed at 49 percent. She sees a fifteen-point gap and draws a reasonable conclusion: pricing flexibility wins work. She tells the partners to be more flexible on fees when the prospect hesitates.

The margin report arrives six weeks later. The firm’s blended contribution margin fell nearly 400 basis points quarter over quarter. Nobody connects the two reports because no report in the firm’s system connects them. The win rate and the margin erosion are the same decision, measured on two dashboards that have never been in the same room.

The pattern is not unusual, and it is not small. BigHand’s 2026 survey of more than 800 legal finance professionals found that 90 percent of firms reported increased client discounts and write-downs, with nearly a third citing average concessions of 11 to 20 percent. That write-off escalation intensified by nearly 40 percent year over year, up from 49 percent expecting an increase in the 2025 report to 88 percent expecting further increases in 2026. Ninety-six percent raised standard hourly rates in the same period. Half of those firms now cite aged work in progress as the primary driver of cash-flow pressure, up from 32 percent the year before. The rates went up. The money collected did not follow.

The number is real. The inference is wrong.

Discounting does improve close rates. The close rate on discounted work is a single number that measures two different things, and the two have opposite economics.

Some of those clients would have closed at full price. The discount landed on work the firm was going to win regardless, and for those matters the concession is a transfer of contribution from the firm to a client who had already decided.

The remaining clients closed only because the discount moved them off the fence. Those are the incremental wins. They are also, as a population, the clients who were comparing firms on price, which means the discount selected for them. Selection is the mechanism, and it governs everything that follows.

This distinction matters quantitatively. Bain’s analysis of dozens of B2B companies across a wide range of sectors found that a 1 percent improvement in realized price produces an 8 percent improvement in operating profit, roughly twice the return from a comparable gain in volume or cost reduction. The arithmetic runs in both directions. Every point of undisciplined discounting erodes profit at the same multiple.

Same line on the dashboard, different money underneath

Two commercial lease matters, both won in the same quarter, both recorded as closed on the partner’s dashboard.

The first was quoted at $180,000 and signed at $180,000. The client chose the firm on reputation, signed in ten days, stayed inside scope, and paid in thirty. Cost to serve ran $108,000. Contribution: $72,000.

The second was quoted at $180,000 and discounted to $153,000 after the prospect mentioned a competing proposal. The competing proposal turned out to be from a solo practitioner the client had no intention of hiring. The client would have signed at full price. But the partner, hearing the word “competitor,” dropped the fee in the room. Cost to serve on this engagement ran $115,000: the client negotiated two rounds of additional review into the original scope and paid in 68 days. Contribution: $38,000.

The partner’s dashboard records two wins at a 100 percent close rate for this matter type. The contribution gap between them is $34,000 per engagement. Nothing in the firm’s reporting surfaces that gap, because the reporting does not know why each client said yes.

The gap is structural, not anecdotal. Clio’s 2025 data puts the average law firm at 38 percent utilization, meaning just 3.0 billable hours captured in an 8-hour day. Median realization lockup is 43 days and median collection lockup is 32 days, for a combined 93 days from work performed to cash received. At $50,000 in monthly revenue, 93-day lockup means $155,000 constantly tied up in unbilled work and unpaid invoices. Discounted clients do not improve these numbers. They make them worse, because they are more likely to dispute scope, delay payment, and compress the margin that was already thin.

Why the blend misleads

Three forces sit inside the blended win rate, each compounding the one before it.

The identification gap. At the moment of proposal, the firm cannot distinguish a client who needs the discount from a client who would have paid the full rate. So the concession is granted broadly, on the strength of a signal as thin as a pause on the phone or a mention of budget. Every hesitation is read as price resistance. Some of it is. A fair portion is the ordinary friction of a purchasing decision, and it would have resolved on its own.

The data on this gap is clear. Bain’s survey of 1,700 B2B executives found that roughly 85 percent of respondents believe their pricing decisions could improve, but only 15 percent had effective pricing tools to support those decisions. Only 13 percent had front-line incentives aligned with pricing strategy. Only 26 percent used dedicated pricing software, despite the fact that companies using such software reported 2.5 times stronger pricing outcomes. The legal industry is worse: fewer than half of firms link matter profitability to partner compensation. BigHand found that 46 percent of firms have introduced remuneration incentives or penalties to reduce write-downs, with another 41 percent planning to do so this year. The partner granting the discount cannot see its margin impact at the time of proposal because no system in the firm connects the proposal to the margin.

Selection. The discount, applied across enough proposals, changes the composition of the client base. Price-sensitive buyers are not a random draw from the market. They negotiate scope after signing. They challenge invoices. They pay slower. They leave at the first rate increase. In subscription commerce, where acquisition channels are tracked more precisely, buyers acquired through heavy discounting churn at two to three times the rate of full-price acquirers. The mechanism is the same in professional services: a discount attracts a buyer profile with lower commitment and lower perceived value of the relationship. Contribution per engagement drops, and cost to serve rises, on the same work, because the client entered the book through a price concession.

The anchor. The discounted rate becomes the client’s reference price. Every subsequent fee conversation starts from the lower number. A rate increase that brings the client back to the original rate card feels, to the client, like a 15 or 20 percent increase, because measured from their entry point, it is exactly that. The introductory discount does not expire. It sets a structural ceiling on the relationship’s lifetime contribution.

The initial concession becomes a permanent drag on the book, renewed with every engagement.

These three interact. The identification gap means the firm discounts broadly. The broad discount selects for price-sensitive clients. The price-sensitive clients anchor on the low number and resist any correction. The initial concession becomes a permanent drag on the book, renewed with every engagement.

The number the reporting cannot show

The measurement that matters is the marginal contribution of discount-dependent wins, net of the margin destroyed on wins that would have closed at full price.

Computing it requires one input the firm almost certainly has and has never assembled: the historical rate at which clients who were initially quoted full price, declined, and then accepted without a discount. That rate gives a floor estimate of the full-price-capable share inside the discounted wins. Subtract the margin surrendered on those. What remains is the contribution generated by the clients the discount actually moved. Divide by the total dollar value of concessions granted across the period.

For most firms, that number is negative. The discount buys incremental volume at a cost exceeding what the volume contributes. The math is not speculative. Bain’s pricing analysis, built on data from more than 1,500 pricing engagements over the past decade, confirms the asymmetry: the profit sensitivity to realized price is roughly double the sensitivity to volume, variable costs, or fixed costs. A 1 percent erosion in price does not reduce profit by 1 percent. It reduces operating profit by roughly 8 percent, because the price concession falls straight through a cost base that does not adjust. The firm does not save money by winning work at a discount. It commits the same resources and collects less.

The firm has been asking whether discounting improves its close rate. That question has a clear answer: yes. The question that governs the decision is different. What is the expected contribution of the client the discount attracted, after subtracting the margin surrendered on the clients who would have paid full price? The close rate was measuring two populations blended into one percentage. One population was profitable at the original price. The other was selected by the lower one.

The discounting policy is the pricing policy

A firm has a rate card and a discounting policy. The managing partner treats the discounting policy as an exception to the rate card, a selective tool for conditions that require it. Run the numbers across a full year and the pattern is structural. Clio’s 2025 data puts the average law firm realization rate at 88 percent. The average collection rate is 93 percent. Multiply them and the firm collects roughly 82 cents on every dollar of standard value. The rate card is the price the firm wishes it charged. The discounting policy is the price it charges.

Thomson Reuters’ 2026 Rates Report confirmed what the math predicts. The report found that firms have self-sorted into three distinct operational models based on rate aggressiveness and tolerance for write-downs. One model sets high rates and maintains strict realization discipline. Another sets moderate rates and offers strategic volume discounts. A third sets aggressive rates and absorbs heavy write-offs to preserve client relationships, what the report describes as firms burning upstream resources the way a jet engine’s afterburner does. Despite being significantly different approaches, all three converged on nearly identical collected rates, between $553 and $580 per hour. The discount strategy did not produce a durable advantage. Market forces, including client expectations, competitive pressures, and economic realities, pulled every model to the same destination.

The 2026 State of the US Legal Market report, published jointly by the Thomson Reuters Institute and Georgetown Law, adds the wider context: the average law firm achieved 13 percent profit growth in 2025, demand surged, and worked rates grew 7.3 percent, more than double inflation. But beneath those numbers, client procurement teams are getting more sophisticated, alternative providers are entering the market, and net spend anticipation among general counsel has dropped to levels not seen since the pandemic. The report calls it a tectonic moment. The firms celebrating record profits are standing on increasingly unstable ground, because the same forces driving today’s success may be setting the stage for tomorrow’s correction.

The win rate confirms the policy is working. The margin explains the cost. Two reports, two systems, one decision nobody made on purpose.

B.L. Sheets is the founder of B.L. Sheets & Co., a firm that delivers Revenue Intelligence and Decision Architecture (RIDA™) for founder-led firms. The Decision Layer is a weekly essay on the revenue, capital, and incentive decisions that hold under constraint.

If you are working through a specific pricing or discounting decision, the Focused Revenue Fit Call is a 30-minute conversation to determine whether a Structural Diagnostic applies.