The Market that Rewards Managing Problems
The dominant revenue cycle business models have a structural incentive not to solve the most fundamental problems in the revenue cycle (denials).
That’s not a criticism of execution. The people building these businesses are sharp, and the services deliver real value. But the architecture of how RCM gets priced (percentage of collections, per-claim fees, FTE-based managed services) ties vendor revenue directly to the existence of adverse payment outcomes. When payers deploy AI to make adjudication more complex and deny claims faster, a traditional RCM vendor’s total addressable market expands.
Systems of Record Never Solve Problems
The SaaS repricing exposed a version of this same dynamic. A large category of companies had been built to capture data, record information and report on it. When public markets stopped rewarding growth at any cost, these companies got scrutinized (and found wanting) on a specific question: what does the software actually do? The answer, for too many of them, was: it records.
CRM platforms organized sales activity, but didn’t close deals. Compliance tools cataloged risk but didn’t prevent violations. Analytics platforms generated reports but didn’t generate decisions. These were systems of record. They were expensive, sticky databases organized around complexity rather than engineered to eliminate it. Customers stayed because the data was trapped, and switching was painful. Not because their problems were being solved.
The companies that survived the repricing weren’t the best recorders. They were the ones that had crossed the line from organizing work to doing it; the ones that became systems of action that automated decisions, surfaced recommendations, and removed problems rather than just logging them.
Revenue Cycle Services Dressed Up as Technology
RCM vendors have run the same architecture for twenty years. They’ve taken on increasingly complex work like utilization management, prior auth, coding optimization and denial resolution. The work is genuinely hard, but the business model means that the harder the work becomes, the more $$ the vendor generates. A payer that tightens medical necessity criteria or deploys AI to increase prior authorization denials isn’t a threat to the RCM vendor’s business; it’s justification for the next contract expansion.
RCM vendors have become highly efficient systems of record for hospital revenue performance. They track denial rates, document appeal outcomes and report collection KPIs. But the root causes of payment problems — documentation gaps created at the point of encounter, DRG inaccuracies assigned at discharge, and auth mismatches — remain upstream clinical problems that post-bill remediation doesn’t reach. Nobody in this system has a strong incentive to ask whether managing these problems is the same as fixing them. For years, with the best of intentions, the market has rewarded the wrong thing.
AI Becomes the System of Action
AI is capable of shifting the intervention point from post-bill remediation to pre-bill prevention, flagging documentation gaps before a physician closes the encounter, validating DRG accuracy before discharge, and aligning authorization before the claim is ever created. This is deployable today for well-defined use cases, and it works without adding work to the clinical team.
That last part matters. The failure mode of most “prevention” technology in healthcare has been that it creates a new operational burden in exchange for downstream savings. What AI makes possible, and what genuinely restructures the market, is upstream intelligence that acts before the problem forms, inside workflows that already exist.
When that capability operates at scale, it doesn’t just improve the process. It restructures the economic logic of the entire market. A platform that prevents the denial doesn’t get paid a percentage of what it recovers. It gets paid on what it protects. The contract structure, value proposition and the multiple all change.
This is the transformation AI is forcing across enterprise software broadly. Moving from systems that record and manage to systems that predict and act. In the revenue cycle, the stakes are specific and quantifiable. Providers are absorbing the financial and operational costs of adverse payment outcomes that increasingly don’t have to exist.
The Revenue Cycle Repricing Won’t Announce Itself
The SaaS repricing was gradual until it wasn’t. The organizations that came out stronger were the ones already asking the harder question. The revenue cycle is at the same inflection point. The market is already shifting from tools that help providers work denials more effectively to solutions that determine whether a denial should have existed at all (proactive denial management).
The vendors and health systems building toward denial prevention rather than denial management are building toward a business model that the market will reward. Those still optimizing the system of record are building toward the same reckoning that SaaS already had.
