🎧 Part I Podcast free on Apple Podcasts and Spotify.
🎧 Part II Podcast episode for paid subscribers only. Also available on Apple Podcasts and Spotify.
To listen to paid episodes in Apple or Spotify, link your Substack subscription via the show settings on those platforms (instructions inside the Substack app under Subscriptions → Podcast).
Table of Contents
The savings headline and the actual market
Provider identity becomes an operating expense
The outsourced investigations business gets more interesting
Risk adjustment starts paying for defensibility
Home care needs a payment clearance service
Equipment suppliers become an underwriting opportunity
Medicaid errors become a managed financial liability
Insurance distribution starts selling proof
Where the durable businesses could emerge
Abstract
The proposed Comprehensive Regulations to Uncover Suspicious Healthcare rule could change what healthcare organizations pay outside companies to do. Provider enrollment would demand more continuous maintenance. Fraud reporting would create additional accountability for how plans investigate and document suspicious activity. Medicare Advantage audit exposure could support businesses built around defensible diagnosis submissions. Home care payment controls could connect verification software with working capital. Equipment supplier disclosures could improve underwriting, while Marketplace enforcement could reward distributors that preserve reliable evidence of authorized enrollment. These are potential business models inferred from the proposal, with different buyers, revenue structures, and implementation risks. The strongest opportunities would earn recurring revenue by completing work that affects whether healthcare organizations can collect and keep their money.
The savings headline and the actual market
Nearly ninety billion dollars in projected taxpayer savings is an excellent way to get a healthcare investor to open an email. It’s also an excellent way to produce a terrible market sizing slide. The Centers for Medicare and Medicaid Services announced the CRUSH proposal on November sixteenth, projecting nearly $90 billion in federal budget savings over the following decade, alongside approximately $378.5 million in estimated costs from 2027 through 2036. Most of the headline economics come from Medicare Advantage risk adjustment audit recoveries, which CMS estimates could average approximately $9 billion per payment year. The underlying analysis treats those recoveries as transfers from health plans to the federal government. That distinction matters to anyone considering a business here: money recovered by Washington does not automatically become money available for software subscriptions. The proposal also remains a proposal. CMS will accept public comments for sixty calendar days following Federal Register publication, with comments due January 19, 2027; final requirements, implementation dates, and subsequent legal challenges will determine how much of the commercial opportunity actually arrives.
The more useful question is what an organization would suddenly have to prove, how frequently it would have to prove it, and what happens financially when the evidence is missing. A provider might need to establish that its ownership disclosures are current. A plan might need to explain how an investigation became a reportable referral. A home care agency might need to reconcile a visit with an authorization before its claim clears. Those requirements create demand for people and systems that finish the underlying work. Plenty of vendors can identify a discrepancy. Fewer can resolve it across a clinical system, a state portal, a billing department, and somebody named Linda who has the only correct spreadsheet.
These businesses would enter markets that already have enrollment specialists, payment integrity contractors, coding companies, and compliance consultants. The opportunity is to change their economics. A periodic project could become a recurring operating service. A software subscription could become a distribution channel for financing. An investigations contractor could charge for maintaining a defensible process even when no recoverable overpayment emerges. The regulatory tailwind would be strongest where failure threatens cash collection or creates a measurable liability. It would be weaker where the new requirement amounts to another form that an existing employee can complete. That’s the commercial test throughout this proposal: identify the operational consequence, then find the budget owner who already feels it.
Provider identity becomes an operating expense
Provider enrollment is a promising starting point because the proposal would make maintaining the record more demanding. It would generally require all Medicare enrollment changes to be reported within thirty calendar days. It also would require eligible providers billing Medicare Advantage to hold approved Original Medicare enrollment, with an exception for the Program of All-Inclusive Care for the Elderly. That proposed requirement has a longer runway, beginning with the plan year starting two years after final publication. Together, these provisions could support an enrollment maintenance business serving providers directly and plans that need their networks ready.
The product would start with a current, documented picture of the organization. It would connect legal entities, practitioners, practice locations, ownership interests, managing personnel, and billing relationships. Each change would generate the appropriate review and filing task. An acquisition closing would prompt an ownership assessment. A location opening would trigger a check of enrollment and contracting requirements before patients arrive. The company selling this service would need access to the customer’s operational changes, because a public database cannot reliably announce that a medical group replaced its managing employee yesterday. Integration with human resources, legal operations, and acquisition workflows could therefore be more valuable than another searchable provider directory.
There’s a technical trap here. A National Provider Identifier does not establish licensure, credentialing, or enrollment, and updating its associated record does not automatically update Medicare enrollment information. Public Medicare enrollment files also have limitations, including the absence of historical enrollment information in the current extract. A serious verification company would preserve its own dated observations and distinguish what the source showed from what was later confirmed. That supports a product capable of answering a useful question: what evidence established that this practitioner, at this location, under this billing arrangement, was eligible on the date the service occurred?
Revenue could combine an implementation fee with recurring charges per practitioner, organization, or location. Plans might buy a network assessment and remediation service, while provider groups might buy ongoing maintenance. A hypothetical customer with ten thousand practitioners paying one hundred fifty dollars annually would generate one and a half million dollars in recurring revenue before additional project work. That’s illustrative pricing, not evidence of willingness to pay. The delivery model would have to make the arithmetic work after document collection, portal submissions, exception handling, and customer support. Acquiring an established enrollment services company could provide an attractive starting point: existing customers and experienced staff, followed by software that reduces the cost of keeping each enrollment current.
The outsourced investigations business gets more interesting
Mandatory reporting could change the revenue model for an outsourced special investigations unit. CMS reports that only forty-six percent of Medicare Part C and Part D contracts voluntarily reported substantiated or suspicious fraud referrals during the reviewed year, while nineteen percent reported payment suspensions based on credible fraud allegations. The proposal would make the referrals mandatory and broaden suspension reporting across provider types. Those figures don’t establish that every silent contract missed fraud. They do explain why the agency wants greater visibility into what plans already discover and how they respond.
A vendor could sell responsibility for moving a case from an initial signal through documented disposition. That includes obtaining records, reconciling claim histories, testing explanations, tracking review decisions, and preparing the material required for an authorized submission. The customer would buy an operating process with named accountability. Pricing could include a recurring fee for coverage, charges for complex investigations, and a carefully defined recovery component. This structure would let the vendor earn revenue for resolving a legitimate case that produces no recoupment. That’s useful because a system paid entirely on recoveries has an awkward relationship with evidence showing that the provider billed correctly. The vendor’s quality measures should include the durability of findings, overturned actions, case completion time, and the amount of unnecessary provider disruption.
Clinical and operational context would determine whether the service deserves its fee. A specialty pharmacy investigation should reconcile product identifiers, quantities, days supplied, paid transactions, reversals, dispensing locations, and prescriber relationships before drawing conclusions. A provider capacity investigation should translate billing units into actual service requirements and account for staffing, operating hours, and the relevant benefit design. High payments alone establish very little. A referral center treating unusually complex patients can look peculiar because its patients are peculiar. Artificial intelligence could help assemble timelines and locate contradictions, but an investigator still needs to distinguish a suspicious arrangement from a messy data feed. An allegation written with polished confidence remains an allegation.
Cross customer learning could eventually improve the economics, provided the company has appropriate rights to use the information. It might learn which patterns routinely disappear after claim reversals are reconciled, which record requests resolve a question quickly, and which findings survive review. It should preserve those outcomes, including cases cleared, rather than training only on successful recoveries. There’s also a market sizing caution buried in the proposal: CMS largely characterizes the reporting work as entry into an existing reporting process. A startup selling form completion alone would have a narrow business. The larger opportunity would come from taking responsibility for the investigative work that makes a submission accurate, timely, and useful.
Risk adjustment starts paying for defensibility
Medicare Advantage offers the largest financial exposure and probably the most crowded sales pitch. CMS proposes extrapolating risk adjustment audit findings beginning with payment year 2018 and proceeding without a fee-for-service adjuster. It is reissuing provisions after a district court vacated similar earlier provisions for procedural reasons. That history deserves a place in any investment model. The proposal expresses the agency’s intended approach; it doesn’t make the future litigation path disappear. A company whose entire value depends on one specific recovery methodology would carry substantial policy concentration risk.
A more durable business would help plans maintain the evidence supporting submitted diagnoses. The operating unit would be a diagnosis linked to its underlying documentation and submission history. The system would preserve the relevant encounter, source record, service date, reviewer decision, and later corrections. CMS audit guidance already addresses record identity, dates, acceptable sources, and other validation requirements. A commercial product could organize that evidence continuously and identify gaps while staff still have a practical chance to resolve them. Waiting until an audit begins can turn an ordinary retrieval problem into an expensive archaeological expedition through systems belonging to a practice that changed ownership twice.
The incentive structure would deserve unusual attention. A vendor paid only for increasing risk adjusted revenue has little commercial reason to prioritize unsupported diagnoses that should be removed. A defensibility service could charge per member, per validated submission, or for a managed review program, with quality measured independently. Its work should include identifying records that support accurate coding, resolving ambiguous submissions, and recommending corrections where appropriate. The output would let a compliance officer trace a submitted diagnosis to a specific record and understand why a reviewer considered it supportable. Language models could locate relevant passages and compare documentation across encounters, while experienced reviewers retain responsibility for judgments that affect payment.
Illustrative economics show why a buyer might care. A plan with one hundred thousand members receiving an assumed average of fifteen hundred dollars monthly would collect about one billion eight hundred million dollars annually. Half a percent of that amount is nine million dollars. That calculation measures the scale of a hypothetical exposure; it does not estimate recoverable overpayments or a vendor’s attainable savings. A service priced at one dollar per member monthly would cost one million two hundred thousand dollars annually. The purchase would still require evidence that the service reduces unsupported submissions or the expense of maintaining defensible records. Before offering any guarantee, the vendor would need to define covered work, establish a baseline, account for older submissions, and determine which outcomes it actually controls.
Home care needs a payment clearance service
Home care could produce a more operationally interesting company. CRUSH would require states to incorporate electronic visit verification information into prepayment review and preliminary investigations. Electronic visit verification itself is already an established requirement for covered Medicaid personal care and home health services involving an in home visit. The proposed change concerns how the information gets used in oversight. That creates an opening for a service that reconciles the visit evidence with the information needed to submit and pay a claim.
An agency could purchase a service that reviews visits before billing and routes exceptions to the person able to resolve them. The system would compare the recorded service with authorization limits, staff assignments, billed units, and relevant state requirements. A mismatch would generate a specific task with supporting evidence. Did the worker forget to record departure? Did an authorization expire? Did a correction arrive after the claim file was created? Treating all three as the same exception would make the software cheap to build and expensive to use. The commercial product would need to preserve original entries, explain amendments, and record who approved them. Providers would pay for fewer avoidable interruptions and less staff time spent reconstructing work that already happened.
Verification design also needs restraint. Federal guidance does not require continuous tracking of a beneficiary throughout the community, and global positioning technology is not universally required by the federal framework. State requirements can differ. A vendor should therefore encode the applicable rules instead of treating every missing coordinate as evidence of a nonexistent visit. It should also distinguish a corrected administrative record from proof that care was delivered. A matching timestamp can support an investigation without resolving service quality, medical necessity, or every coverage condition. The payment clearance service would be most useful when it can explain exactly what has been checked and what remains unresolved.
Financing could become a second source of revenue after the company demonstrates that its reconciliation process predicts collection performance. A lending partner could use the evidence to help underwrite advances on eligible receivables. The software company could earn processing revenue and a disclosed servicing fee, while the lender provides capital under a separately evaluated arrangement. For a hypothetical agency spending two million dollars monthly on payroll, a seven day funding gap represents roughly four hundred seventy thousand dollars of payroll expense. That makes payment timing a concrete purchasing consideration. The financing business would still need to account for payer disputes, recoupments, claim assignment restrictions, and correlated state payment delays. More visit data might improve underwriting. It would not turn a Medicaid receivable into cash already in the bank.
Equipment suppliers become an underwriting opportunity
The proposed equipment supplier provisions point toward another combination of operating software and financial services. CMS would expand disclosures concerning upstream relationships such as marketing, fulfillment, and wholesale arrangements. It also proposes stronger mechanisms for declining bonds from sureties that fail to meet their obligations. These changes could make supplier relationships and surety performance more consequential to continued participation. A commercial response could serve suppliers maintaining their disclosures, counterparties evaluating supplier risk, and sureties deciding which applicants to support.
The useful product would document who participates in obtaining and fulfilling an order, then connect those relationships to the relevant transaction evidence. A supplier might use it to maintain its disclosed counterparties and investigate a problematic marketing source. A wholesaler might use it to assess whether a customer’s business activity is consistent with the orders being placed. A surety might use authorized information to improve monitoring after issuing a bond. The economic value would depend on the decisions supported. A directory of supplier relationships would be easy to replace. A service embedded in onboarding, disclosure maintenance, and transaction review would have more opportunities to earn recurring revenue because customers would depend on it to complete ordinary work.
This could support a verified distribution network in which participating suppliers and upstream companies submit standardized information and maintain it through a shared service. Participation would give a counterparty an organized diligence file, subject to permissions, rather than a vague badge claiming that the business is safe. Revenue could include onboarding fees, annual monitoring subscriptions, and appropriately structured distribution or servicing compensation through licensed partners. The operator would need to preserve independence in how it evaluates participants. Selling a favorable rating to the subject of that rating would be a quick route to destroying the product’s usefulness. The most valuable output would be a documented account of the relationships, evidence reviewed, unresolved issues, and subsequent changes.
An acquisition strategy could provide distribution. An entrepreneur might buy a supplier compliance business, a specialized enrollment operation, or an appropriately licensed insurance intermediary, then add shared data infrastructure. Existing customer relationships would help solve a problem that software startups often underestimate: obtaining current, accurate information from small healthcare businesses. Underwriting benefits would have to be demonstrated through actual outcomes over time. Before expanding into a broader financial product, the operator should know whether its information predicts fewer losses, faster resolution, or lower servicing costs. A compelling network diagram might impress a board meeting. It won’t reimburse a surety.
Medicaid errors become a managed financial liability
Medicaid creates several distinct buyers, and combining them into one sales pitch would obscure the opportunity. Providers face the problem of identifying and resolving their own overpayments. Managed care plans face payment and enrollment reconciliation. States face eligibility oversight, corrective action, and federal financial exposure. CRUSH proposes a Medicaid fee for service reporting and return framework tied to identified overpayments, generally using the later of sixty days after identification or the applicable cost report due date. It would implement an existing statutory obligation through a more explicit regulatory process. A service provider should understand that distinction before advertising the arrival of a brand new repayment duty.
For providers, a managed overpayment service could reconcile remittances, credit balances, duplicate payments, retroactive eligibility changes, and other sources of payment error. Its job would be to move an item through review, quantification, authorized reporting, repayment, and accounting reconciliation. It would preserve the basis for decisions and flag approaching deadlines for responsible personnel. The proposed identification standard includes actual knowledge, reckless disregard, and deliberate ignorance. That makes a product designed to keep management comfortably uninformed a particularly poor idea. The service would need a documented escalation process that brings potentially material issues to the right people while distinguishing an unresolved anomaly from a completed determination.
Pricing could combine a recurring fee for transaction monitoring with project charges for older balances and complicated reviews. Percentage based compensation would require care because the customer is returning money, and the vendor should have no incentive to inflate the amount. The buyer might value fewer unresolved credits, a more accurate liability balance, and less expensive month end reconciliation. Plans could buy a separate service that matches verified member changes with enrollment transactions and resulting financial adjustments. A likely source of differentiation would be the ability to follow an item across systems until the accounting and operational records agree. Sending another exception file to a department already drowning in exception files would accomplish very little.
States present a slower but potentially larger contract opportunity. The proposal would strengthen eligibility and program integrity oversight, while the proposed eligibility audit framework explicitly includes reviews of denied or terminated cases as well as active enrollment. That supports a service concerned with decision accuracy in both directions. A vendor could preserve the evidence behind an eligibility decision, identify the source of recurring errors, and manage corrective work through completion. Payment models should reward validated improvements, including correction of wrongful denials, rather than gross reductions in enrollment. A state should be able to see whether the contractor improved the process or simply made fewer people appear on the rolls. Those are very different results, even when both produce an impressive looking expenditure chart.
Insurance distribution starts selling proof
Marketplace distribution could become a business in which reliable evidence is part of the sales product. CRUSH proposes stronger suspension and termination authority involving abusive broker conduct, along with safeguards against access to Marketplace systems from outside the United States. For an agency, web broker, or carrier distribution team, these provisions could increase the value of knowing who performed an enrollment action and what supported it. A distributor might be willing to pay more for a completed enrollment accompanied by a usable record of authorization and application activity.
A specialist platform could preserve the relevant consumer interaction, consent record, application version, agent identity, and subsequent changes. It should make the evidence retrievable for a complaint or review without requiring staff to assemble it from several vendors. The sales model could charge per active agent, completed application, or monitored member relationship. Larger agencies might purchase the service to supervise affiliated producers. Carriers could require a defined evidence package from distribution partners and pay the vendor to validate its completeness. The platform’s value would come from making authorized work easier to demonstrate and problematic work easier to investigate. A recording that nobody can find during a complaint has roughly the same operational value as a recording that was never made.
Consent evidence needs careful design. A checkbox establishes that a system recorded a checkbox. The more useful record explains what the consumer authorized, which action followed, and whether the person could understand the interaction. Identity controls and version histories would support that record, with access restricted appropriately. Artificial intelligence could assist with locating relevant statements, detecting conflicting instructions, or prioritizing complaints. It should not manufacture confidence from a noisy transcript. A well designed service would preserve uncertainty and send difficult cases for review. That makes it easier for an agency to resolve a complaint fairly and for a carrier to distinguish an isolated documentation problem from a recurring distribution problem.
There could also be a related security service for agencies that need to manage access locations, devices, and agent accounts. Its scope should follow the actual Marketplace access restrictions; a broad claim that every offshore administrative task is prohibited would misstate the proposal. The commercial opportunity would be to make compliant access practical and auditable for legitimate distributors. Over time, verified enrollment quality could influence which agencies win carrier relationships and retain customers. A distributor producing fewer disputed enrollments could justify better economics through demonstrated retention and lower servicing expense. That would give the evidence platform a route into the revenue discussion, beyond a compliance department’s annual request for another small budget increase.
Where the durable businesses could emerge
Across these models, the strongest companies would probably control a specific recurring workflow. An enrollment business would complete updates and resolve exceptions. An investigations company would deliver reviewed case dispositions. A risk adjustment service would maintain evidence supporting submissions. A home care operator would reconcile visits and billing, with financing available where separately justified. Each would need a buyer who can identify the financial or operational consequence of poor execution. Starting with one provider category, one state, or one class of investigation could produce better results than announcing a universal healthcare integrity platform and spending the next two years discovering that every customer means something different by integrity.
The strongest competitive advantage could come from learning which interventions work. An enrollment service could learn which documents and reviews prevent repeated submissions. An investigations firm could learn which signals survive record review and which regularly collapse. A home care service could connect specific exceptions with collection outcomes. Those observations would require contractual data rights, careful access controls, and a disciplined record of outcomes. The proposal should not be treated as permission to obtain otherwise inaccessible claims or distribute confidential allegations. A company’s accumulated knowledge would become valuable when it helps the next customer complete the work more accurately and at lower cost.
Investors should scrutinize the delivery margin as closely as the software demonstration. These businesses may begin with substantial manual work, which can be entirely reasonable if customers pay for a useful result. The question is whether similar cases become cheaper to handle over time, whether integrations can be reused, and whether customer specific rules can be maintained without recreating the service for every account. Financing and insurance revenue should be evaluated separately from software revenue, with the relevant capital requirements and loss exposure included. Recurring income is attractive. Recurring income attached to poorly understood liabilities is less relaxing.
A practical founder could start by acquiring or partnering with an operator that already handles the work, then build software around observed bottlenecks. The early contract might cover a narrow portfolio of enrollment changes, a defined investigation queue, or prebilling review for a specific home care benefit. Expansion would follow demonstrated improvements in completion time, accuracy, and cost. CRUSH could help customers prioritize those purchases, but the company should remain useful if implementation slows or a provision changes. A provider that can keep its enrollment current, support its claims, and resolve payment errors already has reasons to pay for competent help. The proposal could make those reasons harder to postpone.


