Your margins were the secret — and the secret's been leaked. Dentistry is the last profession standing, and here's how it stays that way.
THE WHOLE ARGUMENT IN 60 SECONDS. Medicine sold out. Veterinary sold out — to a candy company. Dentistry is the last clinical profession still majority-independent, and it's slipping: fat margins were the safety net, and the fifty-year real-fee escalator that built them switched off around 2010. Owners — not associates — are absorbing the entire squeeze. But the DSO's only real advantage was the back office, and AI just made a corporate back office rentable by one dentist for the price of a phone bill — at the exact moment the consolidators got trapped by their own debt. When operations equalize, the only thing left to compete on is the chair — and chairside, the independent has never lost. Not every practice can be saved, and most owners won't move. The ones who do get the renaissance. The receipts for every sentence above are below.
This is the most important information you will ignore.
I mean that literally. Most of the people who need this piece will read the title — AI is going to help us? — file it under hype, and go back to the schedule. That reflex is the problem, and it is also the opportunity, because the reflex is evenly distributed and the window is not. So one favor before anything else: notice the reflex. Then read the numbers anyway.
What follows is an argument, not an offer. I have tried to build it the way an argument should be built — a claim, the evidence for it, the strongest case against it, and an honest accounting of what I can and cannot prove. Every figure links to its source. Where the data cuts against me, I've said so rather than routed around it.
I'm writing it because I think this profession is at a hinge, and because the conversation I want to have about it isn't happening anywhere I can find. So I'd rather be argued with than agreed with. If twenty-five years led me somewhere wrong, the people best positioned to show me that are the ones who own the chairs — and if it led me somewhere right, this is worth passing to the colleagues who still believe private practice, and the ethics and dignity that come with it, are worth defending.
One clarification, so it doesn't sit between us while you read: nothing here requires anyone to lose for you to win. This is not a case against the consolidators — they responded rationally to conditions that existed. It's a case about what those conditions just became, and what that means for the last profession in a position to act on it.
And know who's asking. For twenty-five years I have traveled and lived inside dental offices — and inside their finances. Across the desk from the doctor. Standing in the operatories, seeing where the work happens and how it happens. A quarter century of quietly figuring out who's happy and why, who's doing well and why — the pattern behind the practices that thrive and the ones that grind their owners down. Everything in this piece is that quarter century synthesizing at once. Somewhere along the way, thanks to this industry, I raised my kids as a single father — this profession made my small family possible. I know how that sounds, and I know it sounds even worse when a guy leads by admitting it sounds cliché. But the core of me has always wanted two things that don't negotiate: quality time with my kids, and for them to never see the stress I carried as a provider. Every efficiency I ever hunted was in service of those two things. Dental practices paid for that. Dentists' trust made it possible. Working every day alongside people this intelligent has been a blessing I don't take lightly, and I will always be grateful for what this profession gave my family. That debt is exactly why what follows contains no hedge and no maybe. This is what I believe about where dentistry is headed, and I am staking twenty-five years of reputation on it.
Every clinical profession in America has spent thirty years being bought. Medicine is gone: more than three out of four physicians now work for a hospital or a corporation. Veterinary is gone: a candy company is the largest employer of veterinarians in the United States. Dentistry is the last one still standing — roughly seven owners in ten — and the number falls every year.
Here is the thesis, defended below with figures you can check:
Dentistry is the last major clinical profession with a real chance to remain independent, because AI arrived exactly inside the window where a solo practice can still re-arm. Medicine's window closed before the technology existed. Veterinary's closed while everyone was watching medicine. Dentistry's is open now, and it will not stay open.
The mechanics, in five moves:
AI does not kill the private dental practice. AI is the thing that brings it back. That is why this is a renaissance and not a eulogy — a renaissance for the dentists who move, and a clearance sale for the ones who don't.
Line the three professions up, because the pattern only appears in formation.
Medicine. In 2012, 60.1% of physicians practiced privately. By 2024, 42.2% — an eighteen-point collapse in twelve years, by the AMA's own benchmark survey. Count it another way and it is worse: as of January 2024, 77.6% of physicians are employed by hospitals or corporate entities, and for the first time the corporate owners — insurers, private equity, pharmacy chains — passed the hospitals themselves.
Veterinary. Mars — the M&M's company — is the largest veterinary provider in the country, operating roughly 3,000 clinics. Corporate consolidators hold about 25% of general practices and 75% of specialty practices — the high-margin layer went first, as it always does. It got far enough that two U.S. senators opened an investigation into what consolidation is doing to pet care.
Dentistry. Still majority independent — 72.5% ownership in 2023, down from 84.7% in 2005. DSO affiliation has doubled since 2015 to 16.1% of dentists, and among dentists less than ten years out of school it is 27% and climbing. Only 15% of young dentists practice solo, against 48% of the twenty-five-year cohort.
Now sit with what that list means. A candy company examined veterinary medicine and concluded it was a better business than candy — the margins were the best growth it could find. And dentistry? In 2012, the Ontario Teachers' Pension Plan — the retirement fund of Canadian schoolteachers — bought majority control of Heartland Dental, the largest dental support organization in America. Read that again: your profession was purchased as another profession's retirement fund. They rode your margins to a 126% increase in supported practice revenues, then flipped the majority to KKR in 2018 and kept a piece. Schoolteachers in Ontario retired, in part, on the drilling you did in Ohio. Private-equity-backed consolidators are running the identical play through the rest of dentistry right now, and the reason is not complicated: your margins are so attractive that everyone wants a piece of you. The payer takes a piece. The software vendors take a piece. The supply channel, the consultants, the platforms — a piece, a piece, a piece — and now the buyers are circling the whole carcass. You are being picked apart from the outside and the inside, and most of the profession is too busy doing dentistry to notice.
Watch where talented people migrate, too — it's the truest tell in any market. Bankers have left banks to become dental practice brokers and consultants. Accountants have walked away from their firms to advise dentists. I've watched it happen for two decades, and it isn't because they stopped being good at banking or accounting. They ran the numbers and concluded that the best seat in the economy was the one next to a dentist. Smart people holding an unlimited line of credit is a wonderful place to set up shop. That's what juicy margins do: they don't only attract buyers, they grow an entire service layer that lives on you. And every one of those service providers is, right now, running their own quiet AI arithmetic about how much longer their piece lasts.
You should be offended. You went to school. You took the years, the debt, the boards, the hands-on-a-real-person terror of your first extraction. You built something with your own two hands inside your own four walls — and while you were doing the actual work, entire industries organized themselves around getting a piece of the margin you produced. One profession bought you outright as its retirement plan. Be offended. It's the correct response.
Then do the thing that offense is actually for. Here is the difference between you and the physicians and the veterinarians who stood in this exact spot on this exact curve: you can still do something about it — you can control the variables. Not the reimbursement rates, not the DSO across the street, not the cost of gauze. The variables inside your four walls: what your overhead is, who your practice is known by, how a patient feels in your chair, what runs on software instead of salary. That list is longer than it has ever been in the history of this profession, and for the first time you can act on all of it alone. When medicine reached this point, the technology that could have armed the solo doctor did not exist. Dentistry reached the same point in the same decade AI arrived. That is the whole thesis in one sentence: dentists may be the ones AI got to just in time.
And most of you just won't do it. I'll say that three times before this piece is over, because saying it once is a warning and saying it three times is a bet. I'd like to lose the bet.
Something needs saying before the numbers, because I have worked with thousands of you over twenty-five years — the vast majority in the pre-COVID era, face to face, in your offices, not over a screen: being a dentist is a hard job, and nobody warned you. Dental school trained you clinically — superbly — and gave you no business training at all. Then it handed you a small company with payroll, a lease, adversarial payers, and staff dynamics, and wished you luck. The corporate entities circling the profession see none of the humanity in that; they are remote from the day-to-day by design. That remoteness is their business model. It is also, as we'll see, their weakness.
So how did undertrained business operators hold off professionalized capital for three decades, when physicians could not? People credit culture, autonomy, the ADA. The truth is less romantic: the margins. A general dental practice runs 30–40% profit margins. Primary-care medicine runs near 15%, with the all-practice median around 8%. That spread is the entire story. At 8%, one bad payer negotiation ends you, so you sell to the hospital. At 35%, you can overpay for supplies, run a half-empty Thursday, ignore the recall list — and still take home $300K. The margin was the safety net, silently forgiving every business sin nobody taught you to avoid.
The safety net is being cut, strand by strand, and you can watch each cut enter the P&L:
Here is where I stop theorizing, because the profession's own economists have already measured it — and the numbers tell a cleaner story than most dentists have heard.
First, the deepest fact I found anywhere in this research, and it reframes the whole history: for half a century, dentistry rode a real-fee escalator. From 1957 to 2010, dental fees rose about 115% above inflation (BLS CPI series, dental services vs. all items — the cleanest long data in the profession). Twice the purchasing power per procedure, compounding for fifty years. That escalator — not brilliance, not the ADA, not anyone's business acumen — is what built the fat margins, funded the golden age, and forgave every mistake. Around 2010, the escalator stopped. Since 2019, real dental fees have actually fallen behind inflation. The profession is still standing on the wealth the escalator built, and most of it hasn't noticed the stairs quit moving.
Second, the answer to a question worth asking out loud: were dentists always this well paid? No — the golden age had a date on it. ADA's own analysis marks the peak precisely: general-dentist real income peaked in 2005 and has fallen in most years since (ADA HPI, December 2016). Run the multiple against the country: in 1990, the median GP earned about 2.5× the median American household. By the 2005–2010 peak, 3.4×. By 2024: 2.15× — the lowest relative standing in the modern record (ADA medians against Census Table H-5; my arithmetic on two published series).
Then the modern series, in constant 2025 dollars: average GP net income fell from $274,198 in 2010 to $215,320 in 2025 — a 21.5% real decline (ADA HPI, Trends in Dentists' Income, Revenue and Hours Worked, June 2026). And here is the contrast that makes it land: over nearly the same window, real US median household income rose 22.4% (Census P60-286). While the country got a fifth richer, the general dentist got a fifth poorer. A forty-four-point swing in relative position, in one working generation. Physicians, meanwhile, held roughly flat — same decades, and by ADA's own caption: "Dentist incomes are declining while for physicians and surgeons they are stable."
The mechanism is exactly the one described above, and HPI has a name for it: "the fiscal squeeze." Comparing 2016–20 against 2021–25: revenue per dentist rose 1.4% while expenses per dentist rose 8.1%. Reimbursement indexes sit around 129–131 against overall inflation at 138 — while dental staff wages ran to 144. Fees flat, everything else up: the arithmetic of a squeeze, published by the profession's own economists.
And be precise about whose squeeze it is, because this is the finding that should determine how you read everything after this:
Comparing 2011–15 with 2021–25, inflation-adjusted income fell $23,000 for owner GPs — while non-owner incomes were, in HPI's words, "essentially unchanged."
The entire loss landed on the owners. That number is easy to read backwards, so let me be precise about what it means. It does not mean ownership was the mistake. The associate is insulated because they are paid on production; the owner is the one who eats the overhead — every staffing raise, every supply increase, every reimbursement cut lands on their line. Which means the loss is concentrated in exactly the line items an AI back office attacks, and the people bleeding are the only people holding the lever that stops the bleeding.
One more piece of honesty while we're in the data, because it strengthens the thesis rather than weakening it: this is not a demand problem. Dentists per capita have been falling since 2017. Dentists worked more hours in 2023 than in any year on record; ADA's own phrase was "busier than ever." Nominal median incomes are at all-time highs — the compression is entirely a real-dollar, expense-side, owner-side phenomenon. Patients want the dentistry. The margin is leaking out through the cost side of the practice — which happens to be the one side you can now do something about.
Compress a 35% margin from both ends and you arrive at medicine's 15% — and we know precisely what happens there, because the AMA asked the physicians who sold. The number-one reason, cited by 70.8% of them: inadequate payment rates. Margin compression is not a comfort problem. It is the documented mechanism by which a profession loses itself.
And let me describe what the fat-margin era actually permitted, because I watched it happen for twenty-five years and it explains everything about what's coming. A dentist with no business acumen whatsoever — none, and nobody had ever taught them any — could buy a million-dollar practice, make a million dollars' worth of mistakes running it, and still succeed. That was the trial by fire, and the margins were the fireproof suit. Those days are over. Here's the precise shape of what replaces them: the trial by fire gets easier — AI will help you through what used to burn people — but the margin for error goes to nearly nil, because the cushion that used to absorb the mistakes is thinning at the same time. Easier to fly, far less room to crash. The owners sitting on fat margins today can survive almost anything while they re-arm. The people trying to get in are entering a game that no longer forgives.
Which sets up a consequence almost nobody is pricing yet: if getting in from scratch gets harder, existing practices get more valuable — and specifically the part of them that can't be built from scratch. Not the chairs. The goodwill: the name, the recall list, the twenty years of patients who trust one doctor. I'll make the stretch out loud, because I'd rather be early than safe: I can see goodwill allocations climbing toward extremes we'd have laughed at — on a strong fee-for-service practice with a beloved name, something approaching 95/5, goodwill to hard assets — because the market will finally price what was always true: the doctor's name and the patients' trust are the practice. And that raises the stakes on the one event where goodwill lives or dies: the transition. Goodwill transfers through a well-run handover or it evaporates in a botched one — and running that handover well is a craft that takes real experience. In a world where goodwill is 95% of the price, the transition stops being paperwork and becomes the whole ballgame.
But notice what the margin still is, right now, today: funding. The dentist who moves while the cushion exists gets to finance the re-arming out of the cushion. That is the move this entire piece builds toward — and it is a wasting option.
Don't tell me your office is "utilizing AI" if you're utilizing it for something clinical. Every office has a stat like that now — AI in the office: 70%! — and when you look closer, it's the radiograph reader, the scan software, the perio charting. Understand what that is: an efficiency. A tool. Exactly like the tools this profession has been handed before — the CEREC, the intraoral camera, the laser. You watched them arrive, you watched some flop and some stick, you appreciated the keepers, and you were still needed at the end of every one of those evolutions. That is precisely how to look at clinical AI: the next instrument in a long line of instruments. It will make you faster and better, and you will still be the dentist. Don't be afraid on that side. Not even a little.
And if you want proof that clinical technology alone never saves the business, the record already ran the experiment — with clear aligners. In 2002, the aligner pitch to dentists was, verbatim from the manufacturer's own SEC filing, that the product would let doctors "increase patient volume, charge a premium price and reduce chair time." Twenty years later, by the industry's own survey, an aligner case in an orthodontic office billed $5,960 against $6,024 for braces — the premium wasn't just gone, aligners were cheaper — while the manufacturer's revenue per case sat below its 2017 level at nearly three times the volume. And the general dentists who adopted it? The channel's own data shows the average GP shipped about nine or ten cases a year, against an orthodontist's sixty. A product, not a practice. That is what every clinical technology eventually becomes once your competitors have it too: table stakes at a compressed price. Adopt them all — just never confuse adopting an instrument with defending a business.
The business side is the other AI, and it is nothing like an instrument. It is the side this entire piece is about — the back office, the margins, the software, the patient relationship, the discovery layer — and it is the side where you will get crushed if you do nothing. And here is the sad part, the part twenty-five years of watching this profession compels me to say out loud: most of you will do nothing. I have watched how this goes. I've written about exactly how it goes — person by person, silo by silo — and the fuller story is coming to print. The clinical tool you'll adopt without a second thought. The business weapon — the one that decides whether you still own your practice in ten years — most of you will file under "next quarter." That reflex is the whole reason the window exists. It is also the whole reason it closes.
I won't preach what I haven't done. So here's the full confession.
I work independently — essentially a one-man show. I run a practice brokerage with no assistant, no back-office staff, no team. I have never had those things, and for most of my career that was a limitation I managed around. Then AI arrived, and it arrived at exactly the right moment for exactly my kind of operator — because I have spent my whole working life obsessed with finding the most efficient way to do things, and suddenly the most efficient way to do almost everything was sitting there waiting for someone to pick it up.
So I'm going to spill the beans about what I built, and let one real business be the exhibit. My entire back office now runs on an AI system I built myself. I'm not telling you to buy it — I'm not even offering it to you; it's a brokerage back office, it doesn't apply to a dental practice. I'm telling you exactly what it did to my business, because the pattern is what transfers:
| The line item | The old way | The new way |
|---|---|---|
| Email marketing platform | Monthly subscription + hours of building campaigns | Gone. Cancelled. AI writes and sends from my own stack |
| CRM | Big-name platform, four-figure annual bill | Gone. Replaced by one built to fit exactly how I work |
| Listing package assembly | Days of gathering, formatting, retyping | Same day |
| Buyer inquiry → NDA → information out | Next business day, if I wasn't at a showing | Minutes, around the clock |
| Valuation workup | A week of spreadsheet evenings | Hours, more consistent than I ever was |
| First-pass legal and compliance questions | Wait for a callback, bill by the hour | Instant first pass — and the licensed professionals still make the calls that require a license |
| The scheduling/automation glue apps | A ring of small subscriptions | Gone. Line items deleted, not renamed |
| Website — design, construction, every change | Agency invoices, week-long turnarounds, "we'll get to it next sprint" | Built and changed at the speed of thought. I speak the change; it's live. The only bill left is hosting |
(The figures are mine and proportionate to my operation — what a six-figure annual saving means at my scale could mean a great deal more at yours. What matters is the pattern: entire line items disappearing, not shrinking.)
And before anyone reaches for the obvious knife — "he threw caution to the wind, no compliance, get him" — no. Nothing regulated left the licensed human's hands. Every disclosure, every document, every judgment call a license exists for still gets made by the person holding the license. The AI does the processing underneath the accountability, not instead of it. That was the design from day one, because in my business, like yours, the signature is the product.
Here's the part I want you to sit with: people say that those with domain competence — the ones who deeply know their own industry — will be the ones who make the real moves with AI. They are absolutely right, and I'm the proof at miniature scale. I could build this because I've spent twenty-five years inside dental offices and their finances — across the desk, in the operatories, watching who does well and why; the AI supplied the hands, and a quarter century of that told the hands what to do. You have that same domain competence about running a dental practice — more of it than anyone selling you software. That's why this piece exists: one business re-armed is an anecdote. A profession of owner-operators re-arming, each one moving at the speed of thought inside the domain they know cold — that's the comeback. Don't look away. This is the whole concept, running live, in the industry that serves yours.
Then run the exercise every dentist should run. Take the DSO's advantage list — centralized billing, scheduling, recall, marketing, purchasing analytics, reporting — and ask of each line: could one practice deploy this with AI? I will tell you the answer, and I know exactly how it sounds: you could deploy it overnight. I have done it. The things a corporate office assumes are impossible for a single location are, in the current state of the tools, Tuesday-afternoon projects. That is not bravado. That is a field report from someone who bet his own business on it and banked the difference.
And so you know I earned this the hard way: I spent years — and money I don't enjoy remembering — doing it the old way. Freelance developers found on gig platforms, designers in India and Ukraine, talented people I could barely communicate with across time zones and language and briefs that arrived translated into something I never asked for. I was trying to build something genuinely innovative — a way to consolidate how the world looks at dental practices — and the model itself made it impossible. Not because the people were bad, but because I was in the wrong orbit: renting fragments of other people's hours to build something that needed one continuous mind. What I'd been practicing wasn't project management; it was foolishness with invoices. And here's the part that should sound familiar by now — asking your offshore coders whether they're using AI for the coding is exactly like asking the accountant whether the spreadsheet would make a nice replacement for himself. You will get the answer that protects the invoice. The blindness runs deep in this transition, at every layer, and the only fix is to stop asking the people whose billable hours are the question.
The website row above deserves one more beat, because it's the one that still feels like a magic trick to me. My web presence — every site, every page, every change — is now built and altered by speaking. I hold a standing Monday meeting with my AI; it asks me the questions — what happened this week, what's changing, what's working — and it already knows what needs updating before I do. A change that used to be an agency invoice and a two-week turnaround is now a sentence. The only thing I pay for is hosting. I won't pretend the cost of doing business merely "improved" — it dropped in a way I would not have believed two years ago, and the strangest part is how quickly you get used to it.
And I'm not going to lie to you about how I got here, because it wasn't luck. I'm fifty-five. I've been at this twenty-five years — I started with the CDA. The only thing I've ever been "lucky" about is being a guy who was always trying to do two things at once: spend more time with his kids, and find more efficiency in the business to make that possible. Here's the honest footnote to that: no matter what the efficiencies gave me back, I never actually banked the time — the entrepreneur in my head fills every void with the next thing. I can't break away from that, and I've stopped pretending I will. But it's exactly that wiring — decades of hunting efficiencies for the most personal reason there is — that put me this deep into AI this early. And what I want to do with it now is give back to the profession that made my family possible.
Which brings us to the vendors — because what happened to my software stack is coming for the entire dental software industry.
Let me stake four claims in writing, so they carry a date when they come true.
One: the dashboard is dead, and the practice system of the next few years will be a one-trick pony, custom-built for one practice — yours. For years I have watched dentistry's software "evolve," and the latest evolution has been dashboards — dashboards on everything, metrics you were supposed to log in and interpret yourself, as though the software's job ended at displaying your problems in color. That era is over. The next system doesn't show you a dashboard; it speaks to you. You ask it questions and it answers. It finds the issue before you knew to look and tells you about it. And — this is the part nobody's building yet — it knows your end agenda. You will tell it something no practice-management system has ever been told: "I want to treat a large number of patients in my area — enough to live a comfortable life. I want to be ethical, be honest, render the best care I can, get the best CE training available, and be the best dentist I'm capable of being. I want my kids' friends to know my name. I want my kids to be proud that I'm that guy in this community." And it will work toward exactly that — every schedule decision, every recall, every dollar — because that, not a KPI grid, is what running a practice was always actually about. Whatever it cannot do, it will go figure out, and then do. Not a platform you adapt to; software that adapts to you and writes its own missing features. This is not speculation. It is how AI-built software already behaves for those of us using it.
Two: the peripheral products die. All of them. I don't want to scare anyone, but I see this coming plainly, and I'd rather you hear it from someone with no software to sell you: every SaaS product peripheral to your core system is going to be gone. I don't need to name names — name the functions. The one that calls your patients back. The one that manages your reviews. The one that texts the confirmations. The one that does the "AI billing." The one that watches your schedule for holes. Every one of them exists to bridge a gap in technology, and the gap is closing over their feet. One system will absorb them all — call it what it will be: the omnipotent software — and the entire ring of two-hundred-dollar-a-month satellites orbiting your practice goes dark. The moat the incumbent dental software industry stands on is evaporating while they stand on it.
Three: the three-card monte game called SEO is going to go away. Google's game — pay the consultant, chase the algorithm update, guess which card the queen is under this quarter — exists because discovery was mediated by an auction. When patients' own AIs do the choosing — filtered through the patient's real history and needs, cross-checked against what is verifiably true about your practice — the auction dies. You will not "rank." You will be known, or you won't. The winners of that world are the practices that are legible and honest, not the ones with the biggest adwords line — which is precisely backwards from today, and precisely why the change favors you.
Four — and this one I guarantee without qualification: somewhere out there right now is a dentist, way deep in AI, who has already built it. Smarter than me; there are plenty. Right now, tonight, that dentist is baking the omnipotent CRM in a garage — on an AI journey much like my own, just in a different location — and it is already running quietly in one operatory's back office. What that dentist lacks is distribution: the building was the easy part, and getting it into your hands is the only remaining bottleneck. When building is easy and distribution is the bottleneck, the incumbent's runway is measured in months. My money is on the garage. It always should have been.
Which raises the question you should be asking about everyone who advises you on this subject, including me: pick your disciple. There are famous names in this industry who can talk about AI for an hour and twenty minutes without saying one thing that proves they have used it. There are respected advisors polling their programmers on whether the new tools are "ready" — which is the bookkeeping department voting on whether the spreadsheet is ready, while the spreadsheet steps in front of them and says, "I've got it from here."
And learn to hear the defense vocabulary, because it is about to be everywhere: "Hi team!" "Nothing replaces the human touch." "AI can never replace your knowledge and wisdom." Notice who says these things. It is almost never the dentist. It is the vendors, the consultants, the reps — the people billing your practice from five hundred miles away. The human touch they are defending is not the hand on your patient's shoulder; that touch is real, it is yours, and nothing in this piece threatens it. The human touch they are defending is their invoice. When you hear the soft words, that is the tell: you are listening to someone whose piece of you is on the line.
The test cuts through all of it, one question long: don't ask your advisors what they think about AI. Ask what they are using. Then watch half the experts disappear.
Now the part of this almost nobody is saying.
People are using agentic AI wrong. They treat it like a smarter search bar — ask a question, take the answer, close the tab. They are asking it questions instead of asking it to get to know them. That is using a telephone to listen to the dial tone.
The real shape of it is the tin cans on a string. Remember the toy: two cans, a taut string, and the magic was never the audio quality — it was that the line was yours. Private. Direct. One end tied to you, the other tied to something that knew you. And notice the geometry of the toy, because it's the whole point: the two cans are the same size. Almost absolute parity. Neither can is larger, neither is more magnetic, neither end owns the string. For thirty years every channel between a business and a customer has had a fat end and a thin end — the platform's end and yours. The tin cans have no fat end. It is two people communicating in the simplest form there is. Back to basics — which turns out to be the most radical thing the technology does. That is what an AI relationship actually is when it is done right: not an oracle you interrogate, but a companion that accumulates you — your history, your finances, your fears, what you have been putting off and why.
Now let the cans grow into what they actually are: buckets. A bucket of everything that makes you you — what you've done, what you owe, what you're afraid of, how you like to be talked to — and at the other end of the string, your patient's bucket of everything that makes her her. The string between them was always supposed to be the information superhighway. That was the promise, and we all remember it.
Then look at what got built on it instead: toll booths. Every hundred yards, somebody put up a gate. The data brokers who bought and resold your patient list. The ad platforms that charge you to reach people who were already looking for you. The directories that rank you by what you pay. The lead-gen services selling you your own patients back. The listing sites, the review gatekeepers, the "verified" badges. Every one of them a middleman collecting a fee on a road between two people who wanted to find each other anyway — and every one of them selling information that was ours in the first place.
Those toll booths are coming down. Not because anyone regulated them — because when both ends of the string have real intelligence on them, the middle stops being necessary. The road doesn't need a guide when the two travelers can see each other. AI isn't adding another booth to the highway; it is tying your bucket directly to hers, and it is the first technology in thirty years that makes the middle of that road less valuable instead of more.
Here is exactly how it goes, because I've lived the timeline. Sit down with your AI — think of it as a coach — and instead of asking it questions, let it ask you questions. Within one hour, you will have the history of human intelligence on your side. Within a week, you'll have doubts, and questions, and you'll have established boundaries — and inside those boundaries, the beginnings of trust. Within a month, it will lead you — not the way an algorithm leads you, herding you toward whatever pays it, but the way something that actually knows you leads you. Not some algorithm. It knows you. That distinction is the entire era we're walking into.
Why does that matter clinically? Because an AI that knows you can dispel fear precisely because it knows what you are afraid of. Nobody books a therapist to talk about being scared of the dentist. But people will talk to something that already knows them — at midnight, without judgment, without a copay. The barrier to being helped was never the information. It was the confession. A companion that already knows you skips the confession.
I will make the confession myself, since the irony is almost too good. I have walked into thousands of dental offices over twenty-five years; it is my living. And as a kid, I could not sleep the night before a dental appointment — could not sleep — because I was certain I was going to die in that chair. Always the same film: one slip of the drill he is leaning on and it goes into my brain, or straight through the floor of my mouth. And without dating myself too precisely: there were no latex gloves in my dentist's office when I was a boy. Bare hands. It was gross. Nobody ever asked me about any of it, and I would never have told them. That kid needed a tin can on a string.
That kid was not unusual. A census-matched NYU study in JADA found 72.6% of U.S. adults report dental fear — 45.8% moderate, 26.8% severe. Even conservative clinical estimates put real fear at ~36%, extreme fear at 12%, and outright avoidance near 3%. Dentistry's biggest revenue leak was never the fee schedule. It is the patient who doesn't book, and the treatment plan that dies in the parking lot because the patient never really believed it was necessary.
Now follow the string to the chair, and watch the personal bucket protect its person. The patient of the next few years does not arrive cold, and she does not choose cold. Her AI carries her whole bucket — her likes and dislikes, the appointment in 2019 that went badly, the fact that she gags on impressions, that she wants everything explained up front. When she needs a dentist, she does not scroll advertising that caught her in a weak moment. She asks her AI one simple question, and the answer comes back filtered through everything it knows about her and everything it can learn about you. And the night before the visit, it does what no dental office has ever been positioned to do. It leans in and says:
"No big deal tomorrow. Trust me — you're going to be fine at the cleaning. You won't die, Jonathan."
And because it has known you for years and has never once lied to you, you believe it. Pre-visit fear, dispelled. The treatment plan's why, answered before the doctor opens his mouth. Case acceptance, decided before the case presentation.
This is the part worth writing on a wall, because it changes what you should spend money on tomorrow. An AI recommending a dentist to someone it cares about is not counting backlinks and it cannot be bought a better position. It is looking for trust patterns — and trust, to a machine, has a specific shape:
Read that list again and notice what it rewards: being an established, consistent, honest, individually-known practice. That is not a marketing strategy. That is a description of a good private practice that has been sitting in one place doing right by people for twenty years — and for the first time, the discovery layer is about to reward exactly that instead of rewarding whoever spends the most. The dentist who has never been able to out-advertise a corporate competitor has quietly been accumulating the only asset the new system actually counts.
Dentists will appreciate the joke in what these two companions deserve to be called. You have spent your whole career being needled that you are not a "real MD." Fine. In the world that is coming, you employ two MDs, and neither one went to medical school: your Managing Director — the business can, running the operational brain of the practice, defending your margins — and your My Double — the personal can, the one that knows you well enough to finish your sentences, defending you. Every dentist in America is about to have both, or work for somebody who does.
And understand what this architecture ends: the sales-pressure era. Consider honestly what the algorithm does to all of us right now. It catches us at the right time, in the right mood. It knows what we are watching. It picks the moment. It has us nailed — and it works for whoever bought the ad. Your patients live under that machine, and so do you. The AI is the tin can that fights back — the first machine in the history of modern commerce that sits on the individual's side of the string, filtering manipulation instead of delivering it. We will choose what we buy, and which doctor we trust, without the pressure, because our own line answers the algorithm. I believe the correlation history will draw is this: the decline of the independent practice tracked the rise of the middleman machine. Disarm the middleman machine, and the independent comes back — in dentistry first, and then everywhere.
Play the tape forward. AI answers the phone at 2 a.m. It verifies the insurance, fights the claim, fills the schedule, runs the recall, writes the marketing, reads the radiograph, benchmarks the overhead. Both the DSO and the solo office get all of it; the technology does not check how many locations you have.
So what remains to compete on?
By the time the patient is in the chair, there is nothing left but the personal experience. The hand on the shoulder. The doctor who remembers your daughter's name and your fear of needles without consulting a screen. The person whose name is on the door and on the outcome.
That is the one arena where the independent practice has never lost — not once, not anywhere. Consolidation never won on love; ask anyone who has watched a corporate office cycle through four associates in three years which model builds a twenty-year patient relationship. Consolidation won on operations: utilization, purchasing, billing discipline, marketing spend. For thirty years the two arenas were bundled, and operations beat relationship on cost.
And this is not a knock on the people working in corporate offices — it is arithmetic about ownership. Corporate will never replicate the chairside compassion of a dentist who owns the practice, controls the experience, and takes personal pride in every patient who walks out the door — not because corporate employs worse humans, but because pride of ownership cannot be hired, scheduled, or scaled. It is the one input that only exists when the name on the door and the hands in the mouth belong to the same person. The patient experience every AI is now learning to predict and optimize for — calm, continuity, being known — is the experience the owner-operator produces naturally and the org chart produces only by policy. Policy loses that contest every time it's actually measured by a patient who has felt both.
I can say that with unusual confidence, because of a strange privilege of my seat: as a practice broker, I have conducted more DSO exit interviews than the DSOs have. Dentists leaving corporate come through my door from a wider array of situations than any HR department ever sees — associates burned out at year two, veterans who sold in and bought back out, doctors who left quietly and doctors who left loudly — and they all tell versions of the same story. Unless the only dentists a DSO retains are the golden geese — the rare ones who genuinely love patients and thrive at high volume — the patient experience is not the same. It cannot be the same. The people who lived inside both models are unanimous about which room the compassion was in.
Which is exactly why AI ends the commoditization. The consolidation era turned dentistry into a price-and-volume commodity because price was the only signal a confused patient could compare. The tin-can era reverses it: when a patient's own AI understands her values — wants-to-be-known versus wants-it-cheap, continuity versus convenience — it matches her to the practice that fits, and shows her the price is fair within her own value system. People do not abandon their values when choosing a dentist; they abandon them when they can't tell the difference between options. AI makes the difference legible. And people will pay for their value system, every time, once they can finally see it.
And be clear about what people actually do, because we've had it backwards for twenty years: nobody has ever really questioned a price when they could see the value and see that people like them used it and were glad they did. That's not a dental phenomenon, that's human beings. We pay for the surgeon, the mechanic, the contractor our neighbor swears by, without ever asking for the breakdown. Value plus credible social proof has always closed the sale.
So how did dentistry end up haggled over like a used transmission? Because the layer that was supposed to carry the social proof crushed it instead. Your years of training, your thousands of hours of clinical judgment, your artistry — the difference between a margin that seats perfectly and one that doesn't — all of it got flattened into a star rating. One number, sitting next to another practice's number, on a platform whose own revenue comes from selling advertising to the very businesses it ranks. That is not a conspiracy theory; it is simply the business model, in the open. And a five-star scale cannot hold the difference between a dentist who is competent and a dentist who is exceptional, so patients stopped being able to see it — and when people cannot see value, the only thing left to compare is price. That is the commoditization, and it happened to you, not because of anything you did wrong.
Now let me tell you a secret I've learned from sitting inside some of the best practices in this country — the ones whose numbers I've had to examine line by line, which is a privilege almost nobody in this profession gets.
In the very best practices, price is never the conversation. The dentist makes amazing money. The patients love and trust the practice. And cost — the thing the entire industry has organized itself around discounting — is completely overshadowed by trust in the doctor and belief in the treatment plan. Not managed. Not "handled" with financing scripts. Overshadowed. It stops being the subject. I have seen practices with fees well above their zip code's average and a waiting list, and I have seen practices racing everyone to the bottom on price and struggling, sometimes on the same street. The difference was never the fee schedule. It was whether the patient believed the person telling them what they needed.
Which brings me to the line item my industry knows by name and the consolidators have quietly overlooked: goodwill. In every practice sale, goodwill is the part of the value that isn't chairs and cabinetry — it's the relationships, the reputation, the reason patients come back and send their families. DSOs buy on cash flow. They model EBITDA, they model chair utilization, they model synergies — and goodwill mostly shows up on their books as an accounting plug after the deal closes, not as an asset anyone is assigned to grow. Nobody at a regional office has "increase goodwill" on their scorecard. You cannot systematize it, you cannot hire it, and it does not survive associate turnover.
That blind spot was survivable when nothing could measure goodwill. It is about to be very expensive, because the trust layer measures nothing else. What the review economy flattened, the AI era reads at full depth: what you're actually good at, for whom, and why the people it matched with you were glad. The independent's biggest intangible asset just became the most legible signal in the market — and it belongs to the doctor, not the org chart.
You already trust prices this way somewhere — think about Costco. You don't audit Costco's markup; you trust it because like-minded people shop there and the whole arrangement has never lied to you. The same thing is coming to dentistry's clinical side: the patient's AI will explain the crown, the fee, the why, and the trust will ride on matched values the way it rides on that membership card. Algorithmic marketing — the machine that catches people at weak moments — runs in reverse now, and what's left underneath is the oldest arrangement in commerce: the rendering of goods and services between two people whose emotional intelligence and signals matched up.
And to be precise about where you stand in it: AI will help you everywhere, but you are the ultimate sign-off and the fulfillment — and as of now, there is no other way. The hands, the judgment, the license, the pride: still yours, more valuable than ever, because everything around them got honest. However strange the road, this ends up back at the ethics we all wanted at the end of the day — known patients, fair prices, work signed by the person who did it. And no middleman leaning on you for higher production so that an investor on a beach somewhere can have another mai tai — one you bought him with your labor, your hands, your Saturday. That drink comes off the tab in this era. The production pressure was never clinical; it was somebody's cocktail.
AI unbundles them. It commoditizes the entire operational layer — hands every practice, one chair or four hundred, the same tireless back office — and leaves the human relationship as the only differentiator still standing. The consolidator's moat gets filled in. The independent's moat is the only one left holding water.
That is why this is a renaissance and not mere survival: AI is the first technology in the history of the profession that strips away everything about a dental practice except the reason patients loved it.
For as long as anyone reading this has been alive, the Davids of every industry survived on luck and the Goliaths' inattention. That era is ending in the strangest possible way. There is a path — using AI, adopting it fully — where the Goliaths become afraid of the Davids. Not because the Davids got bigger, but because there is no reason to be a Goliath anymore. Scale was the whole point of the giant, and scale just became something a single practice rents for the price of a phone bill. The slingshot is included.
And I'll go further, because I believe it and the record should show it: AI is going to take down a lot of the big corporations, and it is going to happen very quickly. The incumbents cannot move. The infrastructure they spent decades building — the channels, the regional managers, the layers of approval, the processes that made them mighty — is now the hill they die on. Watch the large dental providers stumble over their own org charts while a solo operator redirects an entire back office before lunch. AI moves at the speed of thought, and a corporation, by construction, does not. Speed of thought against speed of committee is not a fair fight, and for the first time in your professional life, the unfair advantage belongs to the small.
You want one hint at how it's actually going inside the big shops right now? They are asking real people to write up their SOPs so their own jobs can be agentified away. Sit with that. The corporation's AI transformation depends on employees documenting themselves out of a paycheck, reviewed by committees, phased across quarters. Tell me how that goes. Meanwhile the fierce hunt for efficiency is easy when the scale is one person — there's no committee, no sabotage, no political layer between the question and the answer. You just move. That asymmetry isn't a gap the incumbents can close with budget, because the budget IS the problem.
Here's the bold claim, and it's about money and timing rather than technology.
The roll-up model was a cheap-debt machine, and the cheap debt is gone. Consolidation didn't accelerate in the 2010s because someone discovered that dentistry needed regional managers. It accelerated because money was nearly free, and the arithmetic of buying practices with borrowed money at low rates was irresistible. Platforms bought at 12–15x EBITDA at the 2021 peak. Then the Federal Reserve moved the funds rate from near zero in March 2022 to 5.25–5.50% by July 2023 and held it there for over a year. The machine's fuel line was cut mid-flight.
Look at what that did, in the industry's own numbers:
Now put the two timelines on top of each other, because this is the part that should make you sit up. The exact years the consolidators spent trapped — unable to exit, unable to refinance cheaply, squeezing existing operations instead of buying new ones — are the exact years AI became capable enough to hand a solo owner a corporate back office for the price of a phone bill. Nobody planned that alignment. It's an accident of monetary policy meeting a technology curve.
But look at the picture it produces. Goliath is in the hospital: over-levered at yesterday's prices, past his exit window, answering to lenders, and structurally unable to move fast enough to use the very technology that could save him. David is in the gym: little or no leverage, no investors to answer to, full executive authority, fat margins still funding him — and a new set of weights that just showed up free of charge.
This is the part I'd underline if I could underline one thing.
AI's entire value proposition to a large organization is removing the middle — the coordination layers, the regional managers, the people whose job is moving information between other people. That is where the savings are. So every consolidator in America is now facing a multi-year project to dismantle a middle they spent fifteen years and enormous money assembling. It has to be mapped, piloted, approved, phased, legally reviewed, and survived politically — by the very people being mapped out of it.
You don't have a middle. You never could afford one. The thing that always looked like your disadvantage — no regional office, no operations department, no analytics team, nobody between you and the work — means there is nothing standing between your decision and your result. The independent doesn't have to cut out the middle. The independent competes where the middle never existed.
Now put a clock on that difference, because the clock is the whole opportunity. A solo owner can stand up an AI back-office function in a week. A large organization needs quarters — realistically years — to do the same thing across locations, against internal resistance, with everyone's incentives pointed the wrong way. That gap is measured in years, and every month of it is a month you are compounding an advantage they cannot yet act on. That is the lead. Not "someday AI might help small practices" — right now, for a measurable window, you can move at a speed they structurally cannot match.
And here's the ending that makes the whole thing almost funny: when the consolidators finally finish removing their middle, what have they become? A practice with low overhead and no bureaucracy — which is to say, they will have spent years and millions turning themselves into you. Except you'll have gotten there first, and you won't owe anyone a return on the journey.
Underneath all of this sits an assumption the entire business world was trained on, and it's quietly expiring. For a generation, the middle-infrastructure era taught everyone the same lesson: problems are solved by throwing money at them. Need capability? Buy headcount, buy platforms, buy the roll-up. And because money was the answer, big money was the big answer — even though deploying big money was always closer to throwing lottery tickets: fund ten platforms, pray two hit, bill the misses to somebody's pension. That is the operating system the consolidators run on. And that era is over — because when capability comes from knowing your domain and directing intelligence, instead of from purchasing layers of people, money stops being the input that wins. A billion dollars cannot buy a DSO the thing this piece is about; a curious owner gets it for a subscription. That is the deepest reason dentistry wins: the profession that could never out-spend anybody just entered the first era where out-spending stopped working.
I don't know how long that window stays open. Rates ease, capital gets patient again, and the machine restarts — the consolidators are not going away, and I'd never bet against capital over a long enough horizon. But right now, for this stretch of quarters, the strong buyer is distracted and the small owner has never been better armed. That is the whole timing argument, and timing is the only part of this you can't get back later.
And now the irony I'd like you to sit with, because it is about you specifically. For your entire career you have had to wear every hat — accountant, marketer, HR department, hiring manager, the person who has to fire someone on a Tuesday and still be gentle with a patient at two. It was too much. It was always too much, and nobody who hasn't done it understands the weight of it.
That burden is now your single greatest advantage.
Because every one of those hats came with something the corporate associate down the street will never have: authority. You can make an executive decision about accounting, marketing, hiring, scheduling, and technology this afternoon, alone, and have it live by Friday — no approval chain, no pilot program, no regional VP. You own a business with margins still fat enough to fund the move, in the last clinical profession where one person controls the whole stack. There has never been a moment in the history of this profession when the solo owner held a bigger structural edge over the giants.
And most of you won't use it.
I'd love to be wrong. That sentence is the reason I wrote all of this — not to be right about it, but to be argued with by a few hundred owners who decide to prove me wrong on their own schedule, in their own practices, starting with one thing this quarter.
Let me put my own record where my argument is. I have competed against national chains for twenty-five years, and on the small scale — inside my own realm — I have beaten them, consistently. Not because I had more resources; I never had more of anything. Because what's inside my realm, I control, and they control almost nothing inside theirs — every decision routed through someone who's never met the client, every relationship handled by whoever's turn it was. The same physics that let a one-man shop out-serve a national brand in my corner of the world is the physics this whole piece is about. I'm not theorizing about the David. I've been him my entire career.
And here's the part of my record I'm proudest of, and the part most relevant to where dentistry is going: I have never sold a thing in twenty-five years in this industry. People say "well, you're selling me right now," and I'd argue back — no, I've survived. I showed up, I was nice to people, I was myself, and it all worked out. Every deal I've ever closed ran on emotional intelligence — reading the room, knowing when a seller needed reassurance and not a spreadsheet, knowing which silence to let sit. That's not a skill everyone has, and it's not a skill most people ever get to exercise — but I found the niches where it's the whole game, and I got twenty-five years of reps. The people who lack it are still selling. And the sellers — in my business and in yours — are the ones who are going to be in trouble, because the one thing the coming era refuses to reward is pressure.
So when I say I can't imagine a future where I don't have a very, very large practice brokerage — that's not bragging, and I want to be precise about why. It's just math. When you compress your cost floor the way I've described and your margins widen, you get to play with your rates in ways your competitors cannot answer. The firm with the lowest honest cost structure and the highest emotional intelligence per deal sets the terms of the market. That math works identically for a dental practice: compress the floor, keep the chairside warmth that can't be hired, and you can price in ways the DSO across the street — carrying its regional office and its investors' returns — structurally cannot match.
If you want to know how this movie ends for the incumbents who see it coming and pass anyway, go ask Yellow Cab how they feel about Uber. They'll tell you — they saw it. They looked directly at the technology, and they refused to adopt it, because the old machine still worked and the new thing was beneath them. This story plays over and over, in every industry, and it never changes. Except one thing is different this time: this time, everyone has the tools. The DSO has them, you have them, the dentist down the street has them. It is no longer a story about who gets access. It's a story about who is willing to reach down, pick them up, and do the easiest part.
It's been said a million times already, and trust me, I'm not nominating myself: when people say there will be an individual billionaire — one person, no employees — they are not kidding. It is going to happen, and it will be someone entrenched with AI who knows how to instruct it, assign work agentically, and deploy at a pace no organization can match. I don't tell you that so you'll chase a billion. I tell you so you'll understand the physics: if one person with AI can outrun a corporation, then one dentist with AI can certainly outrun a DSO's regional office. The same physics, at your scale, is called keeping your practice.
The fair question: if AI hands everyone a free back office, why doesn't independent medicine roar back too?
Argue it honestly, because the honest version is stronger — and it begins with a number that surprises people. Opening a solo primary-care office costs perhaps $70K–$150K. Opening a dental practice costs three to five times that — $250K–$500K. And the gap is not vanity: dentistry is the heavier build. A psychiatrist can practice from a room with a chair. A dental operatory needs water, suction, compressed air, and power run through the slab — the plumbing alone demands more planning than some entire medical offices. On raw physical infrastructure, dentistry is the expensive one.
And understand what that buildout number really means in dentistry, because it defines the psychology of the whole profession: the cost of entry is big, and once the walls are up, the only dependent variable is the doctor's ability to grow the practice. The only one betting on you is you. Corporate dentistry runs the opposite wager — they bet the model can absorb the losers, that the growth machine outpaces the locations they quietly close or lease their way out of. One side bets on a person; the other side bets on averages. Every argument in this piece is about which of those bets AI just changed the odds on.
Which proves the point everyone misses: the wall around independent medicine was never the buildout. It is the system, and the system runs the wrong way:
Dentistry is the inversion. The entire clinical and financial stack — diagnosis, treatment, collection — happens inside four walls one person can own. No admitting privileges. No facility-fee arbitrage against you. Real fee-for-service revenue, patient to practice, no intermediary. The buildout is expensive, but it is financeable expense — practice lenders fund dental acquisitions at high approval rates precisely because 30–40% margins service the debt. A thirty-two-year-old with student loans can still buy the entire machine, plumbing and all, and own every layer of it. A physician cannot buy back independence at any price, because what was taken from medicine was never for sale.
That is the asymmetry the renaissance rests on. In medicine, AI arrives to find nothing independent left to arm. In dentistry, it arrives to find seven owners in ten still holding their own keys — the only profession where handing the individual a free back office changes the outcome.
Every figure links to its source. The bear case is included on purpose.
| # | Statistic | Number | Direction | Source |
|---|---|---|---|---|
| 1 | Physicians in private practice, 2012 → 2024 | 60.1% → 42.2% | Medicine: gone | AMA Benchmark Survey 2024 |
| 2 | Physicians employed by hospitals/corporate entities, Jan 2024 | 77.6% | Medicine: gone | PAI/Avalere 2024 |
| 3 | Corporate ownership of vet practices (general / specialty) | ~25% / ~75% | Veterinary: gone | Brakke Consulting via pets.care |
| 4 | Mars veterinary clinics (largest US provider) | ~3,000 | Veterinary: gone | Yahoo Finance 2024 |
| 5 | Dentists owning their practice, 2005 → 2023 | 84.7% → 72.5% | Standing, eroding | ADA HPI |
| 6 | DSO-affiliated dentists, 2024 (2× since 2015) | 16.1% | Bear | ADA HPI workforce 2025 |
| 7 | DSO affiliation, dentists <10 years out | 27% | Bear | ADA News / HPI |
| 8 | Young dentists solo vs 25+-year dentists solo | 15% vs 48% | Bear | ADA News / HPI |
| 9 | DSO share of dental offices, projected 2026 (offices ≠ dentists; runs ahead of ADA's 16.1%) | ~39% | Bear | Duckett Ladd |
| 10 | Dental practice profit margin | 30–40% | Bull | Overjet 2025 |
| 11 | Medical practice margin (primary care / median) | ~15% / ~8% | Bull, for dentistry | Level CFO |
| 12 | PPO share of dental benefits; reimbursement vs charges | >80%; <50% in many states | Compression is real | Georgia Dental Assn |
| 13 | Physicians who sold citing inadequate payment | 70.8% | The mechanism | AMA 2024 PRP |
| 14 | US adults with dental fear (any / moderate / severe) | 72.6% / 45.8% / 26.8% | Patient-side opportunity | NYU–JADA 2025 |
| 15 | Conservative fear floor (fear / extreme / avoidance) | ~36% / 12% / ~3% | Same, floor case | Cleveland Clinic |
| 16 | Startup cost, dental vs solo primary care | $250–500K vs $70–150K | The moat was never capex | Curve Dental · DoctorsManagement |
| 17 | Share of 2024 acquisitions involving DSO buyers | >55% [NEEDS SOURCE — industry-blog figure (Clerri), no primary dataset traced] | Bear | — |
Editing note: rows 10, 11, and 16 are practitioner-publication benchmarks — directionally solid, widely repeated — worth upgrading to ADA HPI expense-survey and MGMA data before national print.
It goes like this: look at the curve. DSO affiliation doubled in nine years. Twenty-seven percent of young dentists — the future, by definition — are already in. Office-level projections run toward 39%. Medicine's curve looked exactly like this in the early 2000s; everyone said "patients want a doctor who knows them" then, too, and the curve won anyway. Consolidation is a ratchet — every practice sold never comes back, the retiring cohort must sell to somebody, and the DSOs are the best-capitalized somebody at the table. On this reading, AI simply makes the consolidators more efficient, and the "window" is the last daylight before the door shuts.
Three answers.
First, the curve measures the old technology's world. Every data point on the DSO growth curve was earned when scale was the only way to buy a competent back office. That world ended roughly two years ago. Projecting the 2015–2024 curve into 2030 assumes the DSO's core product still costs $2M in payroll to replicate; it now costs a subscription. When the input driving a trend collapses in price by 99%, the trend does not get to be extrapolated.
Second, the young-dentist number is a delay, not a defection — and the cause of the delay is the precise thing that just got solved. I need to correct something I believed until I went looking, because the correction matters more than the original claim. I assumed the collapse in young ownership was a debt story: graduates $300K in the hole, too frightened to borrow more. The ADA's own economists say that's wrong. Their finding, stated plainly: "Educational debt levels do not vary across practice settings for new dentists, suggesting debt is not a major driver of career choice" (ADA HPI, July 2023). Debt is the obvious villain, and the evidence doesn't convict it.
Here is what the evidence does show — ownership by graduating cohort, all measured at the same career point, five to nine years out (ADA HPI, June 2025, n=55,991 dentists):
| Graduated | Owners at 5–9 years out |
|---|---|
| 1996–2000 | 70% |
| 2001–2005 | 67% |
| 2006–2010 | 63% |
| 2011–2015 | 33% |
| 2016–2020 | 21% |
Seventy percent to twenty-one percent in twenty years. But HPI's own conclusion about that cliff is the sentence the bear case never quotes: ownership is delayed, not abandoned — "most roads still, eventually, lead to practice ownership." The 2011–2015 cohort sat at 33% early and reached 58% by ten-to-fourteen years out. They are still coming. They are arriving a decade late.
So what causes the delay? HPI points to work-life-balance preferences and concedes the research is thin. I would add the statistic that stopped me cold in ADEA's senior survey: asked to rate their preparedness across eleven domains, graduating dentists rate themselves 90–98% prepared on every clinical and ethical item — and 62% on "manage a successful business." It is the only item below 80%, and it has barely moved in four years (ADEA, Dentists of Tomorrow 2025).
That is the delay, in one number. Not debt — dread of the business. A generation of superbly trained clinicians who were never taught the operating side and know it about themselves. And what is an AI back office, if not the removal of the exact fear that is postponing ownership by a decade? The DSO's recruiting pitch has always been we will handle the parts you were never taught. That pitch is about to be available to a solo owner for a subscription fee. If ownership stops being frightening at year three instead of year fifteen, 21% is a floor, not a trajectory.
Third — the one I'd bet on: medicine's counterfactual never got to run. The bear case's best evidence is "medicine looked like this and lost." But independent medicine lost without AI ever existing at solo scale — the technology arrived twenty years after the economics broke, into a system with nothing left to arm. Dentistry gets to run the experiment medicine never ran: a majority-independent profession, four-wall economics intact, meeting consolidator-grade infrastructure at solo prices. No profession in history has consolidated after its independents got the same tools as the consolidators — because it has never happened before. This is the first time.
Could the bear still win? Yes — if dentists sit on their hands. The ratchet is real: every practice sold during this window stays sold. That is not an argument against the thesis. That is the deadline.
An argument that only presents its own evidence isn't an argument, it's advertising. So here is the strongest material I found working against what I've written, including the things I went looking to confirm and couldn't.
The one real study on the economics of a dental degree says it's a good investment. The only peer-reviewed ROI analysis of dental education (Stafford et al., J Dent Educ, 2014) found returns averaging 29.4% across the classes of 2003–2011 — roughly double the S&P 500's return over a comparable span — and public-school ROI was rising at the end of their data. Anyone citing that paper for "declining returns," as people do, has to carry the rest of the sentence. Caveat worth knowing: its data end with the class of 2011. Every ROI figure circulating about dentistry today is an extrapolation from fifteen-year-old numbers, mine included.
Applicants are voting the other way. Roughly 12,500 people applied to dental school in 2024, versus about 9,500 in 2004. If the market genuinely believed this degree had stopped paying, that line would bend down. It doesn't.
Debt just improved. Average graduate education debt fell from $312.7K (2024) to $297.8K (2025) — and against cumulative inflation, that's a real decline. The debt spiral I would have described a year ago paused.
And the sharpest one, which points at my own conclusion. The income decline landed on owners (−$26,000) and barely touched associates (−$2,000). Read coldly, that is an argument for being an associate — let someone else absorb the overhead. My answer is that this is exactly backwards for anyone with a ten-year horizon, because the overhead is the part that's about to become cheap, and whoever owns the practice keeps the entire benefit of that repricing. But I want to be honest that the raw number, read alone, argues against me — and that my answer depends on a prediction, while their number is already measured.
Also true: ADA HPI's own 2023 update reported no statistically significant change in GP income that year. Rural practice revenue rose 6.1% while urban fell. Dentists still out-earn nearly every other health profession outside surgery.
None of that overturns the thesis, but all of it should temper it. What I'm claiming is not that dentistry is collapsing — it plainly isn't. It's that the cushion is thinning, the loss is landing on owners specifically, and the tools to reverse it arrived at the same moment the buyers of practices got stuck. If you think that reading is wrong, the case against me is above, and I'd genuinely like to hear the version of it I haven't thought of.
And one more honest thing, because a lifeline is not a lullaby: not every practice can still be saved. Some are too far gone — too leveraged, too tired, too late in the owner's arc — and pretending otherwise would insult the ones who can. The next economic cycle, the first one that arrives without full subsidy underneath it, is going to be a cleansing: small businesses that were barely surviving on fat margins and forgiveness will simply be washed out, in dentistry and everywhere else. I don't say that to frighten anyone. I say it because the difference between the practices that get cleansed and the practices that get reborn is being decided right now, in the quarters before that cycle turns — by who re-arms while the cushion still exists and who spends the cushion standing still.
One conviction underneath everything above.
Everyone building for this industry chases the enterprise contract — the DSO deal, the group-purchasing agreement, the platform play. One contract, ten thousand seats; feeding the demon is how the modern economy works. But stand in an operatory and ask a simple question: of everyone getting paid off this practice — the vendors, the consolidators, the consultants, the platforms — how many have any effect at all on what happens in that chair? Measure the distance however you like — miles, management layers, incentive structures; the unit doesn't matter, because the answer comes out the same. The money has migrated to people whose distance from the chair is total. The hands doing the work and the patient receiving it are together in one room; nearly everyone profiting from the encounter has never seen either one, and nothing they do reaches it.
I am betting the other way. Feed the bottom. Put the supercharger back at ground level, where the doctor and the patient stand in the same room — tools priced for one office, sold to one owner, no enterprise gatekeeper between them. "Save the private practice" is usually a bumper sticker. Priced and distributed this way, it is a business model.
And let's take that phrase back while we're at it, because it was never about nostalgia. Private practice IS the root of this profession. We have all watched it slide, and it has been embarrassing to watch — the one-on-one relationship between a doctor and a patient degrading into a relationship between a consumer and an entity. A brand. A phone tree. A rotating associate whose name the patient never learns. Some people will keep shopping that way; entities will keep existing. But the root was always a person treating a person, known by name, in the same community — and everything in this piece is the case that, for the first time in thirty years, the economics are about to point back toward the root.
And when someone asks whether a solo dentist can really run a business that tight, I tell them about a doctor whose practice I sold. He kept his numbers by hand — 8.5-by-11 sheets, taped end to end, for thirty years. Unrolled across a table it looked like the Dead Sea Scrolls, the paper changing color as the years ran back; at some point we stopped unrolling, because the early years weren't worth risking the artifact. I have called it dental art ever since. But what it really was, was proof: the smartest dentists I have met in twenty-five years know their numbers inside out — not to sell the practice, but to run it. That doctor never needed a DSO. He needed what you now have: a machine that keeps the scroll.
Strip dentistry to its atom: a licensed pair of hands in a mouth for fifty minutes. Hygiene is the recurring event; diagnosis is a byproduct of hygiene; treatment is a byproduct of diagnosis. Everything else — the desk, the phones, the claims, the recall, the treatment talk, the books — is information wrapped around that chair. Information work is what AI eats. The chair is what it cannot.
Phase 1 — strip the wrapper (now to two years). The front office stops being a payroll line. Phones, scheduling, verification, claims, recall, bookkeeping, and the pre-visit fear conversation run on AI for hundreds a month instead of two salaries — ten to fifteen points of overhead returned, zero clinical minutes changed. This phase is pure finance, and it is available today.
Phase 2 — hygiene becomes the diagnostic engine (two to five years). Every recall visit captures everything: AI-read radiographs, scans, auto-charted perio. Diagnosis stops being an event and becomes continuous — which is what it always should have been. The doctor compresses to judgment, consent, and irreversible procedures, confirming across more chairs. Patients arrive pre-calmed; case acceptance rises without a single sales tactic. Revenue tilts toward recurring hygiene membership — dentistry quietly becomes a subscription business with procedures on top. The honest limit: the bottleneck moves to hygienist supply, because chair time is atomic.
Phase 3 — the chair is all that's left (five to ten years). Patients' own AIs choose and book the practice; offices illegible to patient-agents stop existing to them. The endgame inverts consolidation: a two-or-three-op practice at 40–45% overhead, fewer patients a day, more revenue per patient, the doctor earning more while working less — competing on the only thing remaining, the personal experience in the chair.
One caveat spans all three: AI is symmetric. The DSOs get the same tools. The window belongs exclusively to the independents who move during Phase 1, while the margins that kept them alive are still there to fund the move.
Everything above is diagnosis. This is the prescription, and it fits on one page.
Here's how AI brings the private practice back: it hands one owner the back office that used to require a corporation — billing, scheduling, recall, marketing, analytics, the patient conversation — at a price a single office can pay. That was the DSO's entire structural advantage, and it just became rentable.
Here's what you do: audit the layer (every dollar you pay a human or a firm to process information), stand up one AI system this quarter, put your own name back on the front of the practice, start the patient-relationship layer before your competitors know it exists, and know your number.
Here's how you compete: on the only ground that's left when operations equalize — the chair. Pride of ownership, continuity, a doctor who is known. That is the one input a corporation cannot buy, hire, or scale, and the discovery layer is about to start rewarding it.
And if you don't: the margin compression continues, the cushion that would have funded the move gets spent standing still, and the next cycle sorts the profession into the ones who re-armed and the ones who became comps. That's not a threat. That's just the arithmetic on the previous pages.
Not a five-year plan. Monday.
The professions that lost their independence didn't lose it in one dramatic moment. They lost it Monday by Monday — one "we'll look into it next quarter" at a time — until the only rational move left was to sell. Dentistry still has Mondays left. Not many. But it is the only clinical profession in America that has any at all, and for the first time in thirty years, the technology is on the independent's side of the string.
Pick up the can.
And then take the challenge, because it costs nothing but a minute. Tell me I'm wrong — publicly, privately, a one-line reply, a comment, a post of your own. If twenty-five years led me somewhere wrong, say so. If it led me somewhere right, say that louder. Then send it to one dentist. Just one. The classmate you still text, the colleague you argue with at the study club, the friend two towns over quietly wondering whether to sell. There is no association, no vendor, and no platform whose interests are served by dentists working this out together — which is exactly why it has to travel the only way anything real in this profession has ever traveled: one owner telling another. A few hundred of us arguing about this out loud is a movement. Nobody forwarding it makes it a long letter from a broker who cared.
So: send this to somebody who still believes in private practice. A classmate, a study-club friend, the colleague two towns over you argue with at meetings. This conversation only works from the bottom up — one owner to another, no platform in between — which, if you've been paying attention, is the entire point of the piece.
Numbers talk better than arguments, so here is the whole thesis as one grid. An illustrative general practice at $1,000,000 in collections — categories and starting percentages drawn from the industry averages cited throughout this piece, rounded for discussion.*
| Expense silo | Before (typical today) | After (AI back office) | What changed |
|---|---|---|---|
| Clinical payroll (hygiene, assisting) | 25.0% | 24.0% | Almost nothing — the hands stay human† |
| Front office / admin payroll | 8.0% | 2.0% | Phones, scheduling, verification, recall → AI† |
| Billing & collections services | 3.0% | 0.5% | Claims chased by software that never sleeps |
| Software / SaaS stack | 2.5% | 0.8% | The satellite ring → one system |
| Marketing & advertising | 4.0% | 1.5% | The ad auction dies; legibility is nearly free |
| Consultants | 1.5% | 0.2% | The binder is a same-day AI output |
| Supplies & labs | 13.0% | 12.5% | AI-watched ordering, benchmarked pricing |
| Occupancy | 8.0% | 8.0% | Rent is rent |
| Insurance & other overhead | 5.0% | 4.8% | Mostly structural |
| Total overhead | 70.0% | 54.3% | ≈16 points recovered |
| NEW SILO — Patient & Employee Appreciation | 0% | 3.0% | see below |
| Owner margin | 30.0% | 42.7% |
Now read the grid again for what it actually proves. Every line that collapsed was a middleman standing between you and your own information — the billing service that sat between you and your money, the marketing agency between you and your patients, the consultant between you and your numbers, the software vendors between you and your own practice data. That was the whole business model of the middle-infrastructure era: park yourself in the gap, charge rent on the crossing. AI didn't make those companies worse at their jobs. It closed the gaps they were standing in.
Which flips the direction that everything has run for thirty years. Your practice has been shaped from the outside in — payers setting your fees, platforms setting your visibility, vendors setting your systems, consolidators setting your exit. What the grid describes is the reversal: from the inside out, powered by you, the doctor. The information starts with you, the decisions start with you, the margin stays with you, and the layer that used to convert your work into their revenue simply isn't in the road anymore.
Look hard at the new line, too, because it's my favorite one on the grid. When sixteen points come back, you don't have to keep them all. Roll a few forward — profit sharing, team bonuses tied to practice goals, patient appreciation, the birthday lunches, the fun — a silo that has never existed on a dental P&L because there was never room for it. Attach it to revenue, let the team see the connection, and the practice stops being a place people work and becomes a mission people share. The most defensible moat in this entire piece — the chairside experience — is produced by people who feel the winning. Now there's a budget line that pays for it.
*Illustrative model, not a promise: every practice's numbers differ, and your starting percentages may sit above or below these. Run your own layer audit (Monday Morning, item 1) before believing any grid — including mine.
†And the honest footnote the grid demands: this table does not show the replacement of the human element. I acknowledge that. Front-office roles in this model shrink, and those are real people — some of whom have been with you for years. Some will be redeployed into higher-value work: patient experience, treatment coordination — the appreciation silo exists partly to fund exactly that — and some transitions will be genuinely hard. Pretending otherwise would break the trust this piece runs on. The wave arrives whether we acknowledge it or not; I'd rather you steer it with your eyes open than have it steer you.
P.S. — a standing offer, not a flourish. I'm not selling anything here; I want to hand you something.
I can tell what kind of business person a dentist is from a single document. Send me your lease. I will write you back and tell you what you signed, why you signed it, and what it says about how you run your business. No charge, no pitch. It is the fastest honest mirror in this industry — and if this piece made you wince even once, you already know what the lease is going to say.
— Jonathan Ingalls. Companion documents: THE-DENTAL-THESIS.md (operator doctrine, internal), THE-CHOKE-PERSON-MEMO-FINAL.md (the accountability argument), YOUR-AI-VS-THEIR-ALGORITHM-THESIS.md (the tin-cans doctrine at length).
© 2026 Jonathan Ingalls · tdibroker.com · Argue with me: jonathan@tdibroker.com