After the roll-up: what provider platforms must now prove, practice by practice

From buying growth to building it

For a decade, the provider platform playbook rested on acquisition: buy independent practices at a modest multiple, combine them, and sell the combined earnings at a higher one. That playbook is running out of room. Physician practice deals fell from a peak of 851 in 2021 to 105 in the first half of 2026, according to PitchBook data, and McKinsey found that private equity investment in physician assets fell by a third in 2025.

For a sponsor or a CEO, the implication is direct. With fewer add-ons to buy, value has to come from the practices already owned, and buyers will want to see it practice by practice.

Several forces are slowing roll-ups at once

Bain points to labor shortages and reimbursement pressure from public and private payers. States are adding oversight. Several now review health care transactions, and Oregon has gone furthest: its 2025 law bars management services organizations from controlling physician hiring, schedules, compensation, coding, billing policies, and payer contracts, for new arrangements since January 2026 and for existing ones from 2029.

Deals still happen, and dental, dermatology, musculoskeletal, and vision practices remain active. But McKinsey finds sponsors favoring targeted add-ons, adjacencies, and operating leverage over broad new platforms.

A track record of organic growth

Bain’s 2026 health care private equity report names the new exit priorities: a track record of organic growth, repeatable playbooks for ancillaries and acquisitions, and a reputation as an employer of choice for clinicians. The buyer pool has also widened. Distributors such as Cardinal Health, Cencora, and McKesson have bought specialty platforms, and strategic buyers like these will examine organic performance closely.

A platform that cannot separate same-store growth from acquired growth will struggle to defend its value.

Physicians leave after the exit

The fourth question deserves particular attention. A study by Berquist, Klarnet, and Dafny in JAMA Health Forum followed physicians at practices that private equity owners sold between 2016 and 2018. Two years after the sale, those physicians were 16.5 percentage points less likely to remain at the practice than comparable physicians elsewhere, a drop of more than a quarter in retention. They did not retire; they moved, often to larger groups. The authors suggest that retention incentives from the first deal expire at exit, and that the next owner may push harder on margins.

For a platform heading toward a sale, physician stability is a leading indicator to track now, not a surprise to discover in diligence.

Judge it at the practice level

AI tools for scheduling, intake, documentation, and revenue cycle are arriving in practices quickly, but they do not pay off equally everywhere. Tools that guide front-desk staff through intake and scheduling tend to help most where management is thin or turnover is high, which describes many recently acquired practices. Practices still being integrated generally benefit from guiding and offloading routine work before they attempt to redesign whole processes. In every case, the test is whether a tool moves capacity, show rates, or collections at the practice level, not whether it is in use.

None of these answers comes from a standard report

Each acquired practice arrives with its own practice management system, its own definitions of a visit, a new patient, and a provider, and its own approach to coding and billing. The first step is a governed data foundation: a semantic layer and a shared model of practices, providers, patients, and appointments that holds across every acquisition.

The second step is economics. That means decomposing growth into same-store, acquired, de novo, and mix components; estimating what no-shows and unfilled slots cost by site; and modeling which providers are at risk of leaving and what each departure would do to revenue. A dashboard is only the last step, the place where that work becomes visible to the people who act on it.

The same foundation also changes who can do the analysis. With clear data governance in place, defining who owns each data domain, who stewards its definitions and quality, and who maintains the systems as custodian, practice and regional leaders no longer have to queue every question behind IT. A well-built semantic layer, AI tools that work against it, and the right training let operations, revenue cycle, and finance teams build the analytics they need themselves, while IT concentrates on keeping systems reliable and available.