By Theo Sakalidis, Co-Founder & CEO, Cevi
LinkedIn: Theo Sakalidis
LinkedIn: Cevi
When I worked on the buy side of healthcare private equity, we underwrote a lot of physician practices. And I can tell you that the clinical work, the thing the doctors trained a decade to do, was almost never the thesis. The thesis was the back office.
That sounds cynical. It was just the math. When you model a roll-up, you are not betting that the cardiologists or dermatologists will suddenly practice better medicine under new ownership. You are betting that you can take twenty or thirty practices, strip the administrative function out of each one, and run it once, centrally, at a scale none of them could reach alone. Billing, eligibility, prior authorization, denials, collections, patient access, collapse all of it into a single shared-services operation with payer-specific playbooks and the leverage to negotiate, and you have manufactured margin that did not exist before. The clinical revenue was the reason to show up. The administrative arbitrage was the return.
I want to be precise about what that moat actually was, because the entire logic of physician consolidation rests on it. The moat was one durable assumption: that administrative competence requires headcount, and headcount requires scale. For decades that assumption held without exception. A four-physician group physically cannot staff a denials-analytics function. It cannot afford a dedicated prior-authorization coordinator who knows each payer’s quirks cold. It cannot build the reporting to see, across thousands of claims, which diagnosis-procedure combinations its largest payer quietly started rejecting last quarter. So it does the only things available to it. It absorbs the administrative tax as lost margin, or it sells, or it affiliates, and slowly the math makes selling look less like surrender and more like relief.
This is not a fringe dynamic. A Government Accountability Office review last year cited AMA data showing roughly 47% of physicians were employed by or affiliated with hospital systems in 2024, up from under 30% in 2012, with the share at private-equity-owned practices rising from 4.5% to 6.5% in just two years. And when you ask why owners take the call, the answer is rarely “I wanted to practice worse medicine.” Administrative burden and staffing pressure consistently rank among the strongest forces pushing physicians toward employment. The paperwork is the push. Consolidation is the landing spot.
Here is the part that should make every acquirer slightly uncomfortable, and every independent owner pay close attention. The assumption underneath the moat is breaking. Administrative competence no longer requires headcount, because the functions that justified centralization are exactly the ones now automatable at small scale. Eligibility verification, denial-risk scoring before a claim goes out, authorization status tracking, statement and collections workflows. These are high-volume, rule-bound, measurable tasks, which is to say the tasks software is genuinely good at. A three-physician practice can now stand up an operation that, on the metrics a payer actually sees (clean-claim rate, days in A/R, denial-overturn rate), looks like a forty-provider platform’s. The thing that used to be rolled up is becoming something you can rent by the month.
Do not take my word for it. Take the buyers’ word for it. Bain’s read on the current private-equity environment is that investors are now leaning on AI-enabled revenue cycle management and back-office automation specifically to cut the administrative work that drives physician burnout. Read that carefully, because it is an admission against interest. The smartest capital in the space is conceding that the administrative moat is now a software problem rather than a scale problem. And the moment it becomes a software problem, the independent practice can buy the same software. The arbitrage compresses. The gap I used to price for a living gets narrower every quarter.
Three forces are converging in 2026 to sharpen all of this.
First, the regulatory wind has shifted against the roll-up just as its economics wobble. California’s AB 1415 took effect on January 1, expanding state notice requirements for transactions involving management services organizations and private-equity funds. The American College of Physicians published a May 2026 position paper calling for tighter oversight of private equity to protect clinical autonomy. And a joint FTC, DOJ, and HHS inquiry into private-equity control of care signals this is a federal priority, not a state-by-state curiosity. Deals are getting slower, more expensive, and harder to defend.
Second, the administrative work that most rewarded scale is being turned into structured data, and nowhere faster than in prior authorization. The 2024 CMS Interoperability and Prior Authorization rule put FHIR-based authorization APIs and firm decision deadlines in place, and in 2025 roughly 80% of the insurance industry pledged to eliminate prior authorization for common services such as imaging, physical therapy, and outpatient surgery. The reason prior authorization rewarded scale was that working fragmented payer portals and fax queues took staff. When that work becomes an API call and a structured response, headcount stops being the constraint, and a two-person practice can run what used to require a department.
Third, the recoverable prize is large, and the playbook for capturing it is already proven. The 2025 CAQH Index estimates that U.S. healthcare avoided roughly $258 billion in administrative costs in 2024 by shifting transactions from manual to electronic, with another $20 billion in annual savings still on the table from the transactions that remain manual. Look at who is sitting on that remaining opportunity. It is not the integrated systems that automated years ago. It is the smaller practices still verifying eligibility by phone and chasing claim status by fax. The administrative savings that scaled platforms captured a decade ago are finally within reach of the practice that never had the headcount to capture them. That is the thesis in one statistic. The back-office advantage was always a function of who could afford to automate, and the price of automating just fell through the floor.
I want to be honest about the limits of this argument, because the maximalist version of it, the one that says AI saves independent medicine and the roll-up is dead, is wrong, and you should distrust anyone selling it.
The moat that is draining is the administrative one. The others are still standing. Consolidation still buys real leverage with payers, and no eligibility tool fixes a bad contract. It still brings recruiting power and capital that a solo owner cannot match on cash flow alone. AI narrows one specific advantage. It does not erase the case for scale.
It is worth being just as honest about when partnering is the right move, because often it is. A senior owner approaching retirement with no partner ready to buy in gets something no software provides: a sale turns a career’s worth of equity into liquidity and gives staff and patients continuity. And when what you actually need is capital rather than relief from paperwork, a partner with a real balance sheet is not a crutch, it is the only way the project gets built. Adding an ambulatory surgery center means real estate, an operating-room build-out, equipment, and licensure that organic cash flow rarely covers. A young fertility physician opening a de novo practice needs an embryology lab before the first cycle: clean rooms, cryostorage, incubators, and the air handling and quality controls that run into the millions. Radiation oncology needs the linear accelerator. Advanced imaging needs the MRI suite. None of those are administrative problems, and not one of them is solved by an eligibility bot. That is exactly the point. The reasons to consolidate that turn on capital and facilities are as strong as they ever were. The reason that turned on drowning in administrative work is the one quietly dissolving.
And independent practices have been promised software salvation before. The EHR era was sold as efficiency and delivered, for many, years of documentation misery. The owners who actually benefit from this shift will not be the ones who buy the most AI. They will be the ones who pick a single administrative function, usually the one driving their largest denial category, automate it with a human still reviewing the exceptions, prove the return over a quarter or two, and only then add the next. Bought carelessly, this technology becomes one more expensive layer solving a problem you did not have.
So here is what I would tell an owner weighing the call from the platform, with the candor of someone who used to make that call.
Re-underwrite your own independence before you sign anything. If administrative burden is the main reason you are considering selling, recognize that it is the one reason with a rapidly shrinking half-life. The offer in front of you was priced against last year’s automation economics.
Then audit your administrative costs honestly. Separate what is genuinely scale-driven (rate leverage, capital) from what is now software-addressable: eligibility, denial rework, authorization tracking, patient statements. Put a real number on what those functions cost you in staff hours today, and what they would cost automated. That delta is the moat. It is the value a buyer is counting on extracting from you. The question worth sitting with in 2026 is no longer simply whether to sell. It is whether the thing a buyer is paying you for is something you could now, for the first time, keep for yourself.
I spent years on the side of the table that profited from the answer being no. I think the answer is finally starting to change.