The following is a guest article by Megan Schmidt, President Chief Executive Officer, Madaket
Health plans are pouring capital into automation, analytics, and artificial intelligence, betting that smarter systems will lower administrative costs and improve operational performance. Yet a growing body of evidence suggests many of those investments are stalling, not because the technology is immature, but because the underlying provider data is fundamentally unreliable.
The situation has been described as a “house of cards,” with inconsistent and outdated provider data preventing health plan AI initiatives from ever reaching scale. The reporting reinforces a basic truth that has long haunted enterprise technology: AI systems cannot outperform the quality of the data they rely on.
What has received far less attention, however, is the financial implication.
For payer CFOs, unreliable provider data is not simply an IT challenge. It is a persistent, compounding cost center that quietly inflates administrative expenses, drives claims rework, and undermines the return on digital transformation investments.
The Invisible Cost Behind Provider Data Chaos
Every health plan maintains provider information across a sprawling ecosystem of systems, including enrollment, credentialing, directories, claims, network management, delegated entities, and external vendors. When a provider updates a specialty, location, tax ID, or affiliation, that change often must be verified and re-entered multiple times across disconnected platforms.
This work rarely appears as a discrete line item on a budget. Instead, it is embedded across departments and absorbed as routine operations. Teams spend time reconciling conflicting records, correcting downstream errors, responding to directory inaccuracies, and managing avoidable provider and member disputes.
Because these costs are distributed, they are rarely measured directly, even though they recur continuously. For CFOs under pressure to reduce SG&A while supporting growth initiatives, provider data represents a blind spot with real margin impact.
Why AI Is Exposing the Problem, Not Solving It
AI initiatives are forcing health plans to confront a long-standing operational reality: provider data is not standardized, synchronized, or authoritative.
Machine learning models can optimize workflows, flag anomalies, and automate decisions, but they cannot compensate for conflicting versions of provider truth. When underlying data varies across systems, AI outputs become unreliable or misleading. In practice, many organizations find that automation initiatives stall in pilot mode, staff revert to manual validation to double-check results, and new tools are layered on top of broken data foundations.
From a finance perspective, this dynamic is particularly painful. Capital is deployed, vendors are onboarded, and internal teams are mobilized, yet the expected efficiency gains fail to materialize. The issue is not that AI does not work. It is that the economics collapse when the data feeding it is fragmented.
The CFO Angle: Provider Data as an “Admin Tax”
Most payer CFOs can track medical cost trends, utilization, and claims performance in granular detail. Far fewer can quantify how much their organization spends each year maintaining provider data accuracy.
Yet poor provider data drives tangible financial consequences that directly affect margins:
- Preventable claims rework and denials that increase administrative expense
- Delayed provider onboarding that slows revenue realization and network growth
- Directory inaccuracies that create compliance exposure and reputational risk
- Labor costs that scale with volume because manual correction never disappears
In effect, fragmented provider data functions as an “admin tax,” a recurring operational expense that grows as plans add members, expand networks, or acquire new entities. Unlike medical costs, however, this tax is not inevitable.
From Fixing Data to Fixing the System
Industry attention is beginning to shift away from point solutions that clean up provider data after problems surface and toward infrastructure approaches designed to prevent fragmentation at the source.
This emerging model treats provider data less like static records and more like a continuously updated supply chain, where validated updates are normalized once and automatically synchronized across participating systems. Instead of multiple teams chasing the same information, organizations operate from a shared, authoritative provider record that stays current over time.
For CFOs, the appeal is straightforward. Fewer manual touches reduce operating expense. Auditability improves. Compliance confidence increases. Most importantly, the infrastructure scales without requiring proportional headcount growth.
Just as critically, this approach restores the economics of AI by giving automation a stable and trustworthy foundation on which to operate.
Why This Matters Now
With margins tightening, regulatory scrutiny increasing, and AI budgets under closer examination, health plans are being forced to justify not just innovation, but measurable outcomes.
Provider data may not be the most visible challenge in payer operations. But as AI initiatives continue to expose its weaknesses, it is becoming increasingly clear that fragmented provider data is not a technical inconvenience.
It is a financial liability hiding in plain sight.
For CFOs looking for sustainable efficiency gains, the question may no longer be whether to invest in AI, but whether their provider data infrastructure is capable of supporting it at all.
Megan Schmidt is based out of Grand Rapids, Michigan, and is the President and Chief Executive Officer at Madaket. With deep experience at some of the nation’s largest integrated care delivery systems, Schmidt previously held senior executive leadership roles at HealthPartners and Corewell Health, which includes healthcare providers and its own health plan, Priority Health.
No comments:
Post a Comment