The cost squeeze on medical practices is no longer anecdotal. A June 24 MGMA Stat poll found 84 percent of medical groups report year-to-date operating costs higher than the same period in 2025, with an average increase of about 11 percent across 251 responding organizations. Meanwhile revenue growth has not kept pace - earlier MGMA polling through the spring showed a shrinking share of groups reporting revenue gains.
The timing compounds the pressure. Practices absorbed several years of labor inflation, and 2026 adds payer-side friction on top: more prior authorizations to chase, more documentation requests to answer, more denials to rework. Each of those is a task that lands on the same staff whose hours are already the fastest-growing line item. The squeeze is not two separate problems - cost and workload are the same problem seen from the CFO's desk and the billing office floor.
An 11 percent cost increase does not get solved by a coding tweak. It forces a structural question: which operating expenses can be reduced without touching patient care or collection performance? For most practices, the honest list is short, and near the top of it sits revenue cycle administration - the hold time, the portal logins, the denial rework.
Why RCM costs resist traditional cuts
Billing is usually one of the largest non-clinical line items, but it has proved hard to cut with conventional tools:
- Headcount cuts hurt collections. Fewer billers means slower follow-up, which means higher days in AR and more denials aging past recovery. The savings show up in payroll and the losses show up in cash.
- Outsourcing trades margin for distance. Vendor costs are predictable, but visibility into claim-level work often drops, and the practice still owns the payer relationship problems.
- The work itself is unbounded. Payer portals multiply, rules change, and every added payer or plan design adds manual steps. Manual RCM costs grow with payer complexity, automatically.
That last point is the trap. A billing team sized for last year's payer behavior is undersized for this year's by default, because the payer side adds friction faster than practices add staff. The cost is also invisible in unit terms: no single portal check or hold session looks expensive, but multiplied across dozens of payers and hundreds of claims a week, it becomes one of the largest controllable line items in the practice.
What actually bends the curve
The practices that reduce RCM cost without sacrificing collections share a pattern: they remove labor from the repetitive layer, not the judgment layer.
- Automate the status check. Claim status inquiries are the highest-volume, lowest-judgment task in billing. Taking them off the phones eliminates hours per day per biller.
- Automate denial intake and routing. Reading remittances, classifying denial reasons, and starting the right workflow does not need a human on every case - only on the exceptions.
- Make the exception queue the job. When routine follow-up runs itself, experienced billers spend their time on the accounts that actually need judgment: complex appeals, COB tangles, high-value claims.
A useful way to scope the opportunity is to tally where biller hours actually go in a typical week: status checks, portal logins, hold time, denial triage, resubmissions, and appeals, in roughly that order. The first three are almost entirely mechanical. If they consume half the team's hours - and in most shops they do - then the automation math is straightforward: those hours convert to capacity without a single new hire, and the judgment work that actually recovers dollars finally gets the attention it deserves.
This is the arithmetic that makes AI claim follow-up a cost story rather than a novelty: the routine layer is the majority of the labor, and it is the part payers have already automated on their side. The alternative - holding headcount flat while payer complexity rises - is not actually flat; it is a slow decline in follow-up quality that surfaces later as higher AR and worse denial recovery.
The takeaway
An 11 percent cost increase with flat revenue is a mandate to re-examine how the back office works. Cutting billing staff is the false economy; the collection losses arrive on a delay and get attributed to something else. The durable lever is removing the repetitive work - status checks, denial triage, follow-up chasers - so the same team can absorb more payer complexity without growing payroll. That is the division of labor QHB is built on: agents work the queue, your team works the exceptions.