ClockSort — credit-balance sorter for medical practices

1. One-liner

Tells a practice which patient credits face a 30-day, 60-day or escheat clock — and which to leave alone.

2. Trend signal — why now?

A patient credit balance is the most over-regulated dollar in a small practice, and as of January 2026 it sits under three clocks that disagree with each other.

1. Florida CS/CS/SB 1808 took effect 1 January 2026 and imposes a 30-day refund clock from “determination” of a patient overpayment. It reaches both facilities (§408.032(8)) and individually licensed practitioners (§456.001) — physicians, dentists, nurses, pharmacists, therapists and 20+ other professions. Facilities face administrative fines up to $500 per violation under §408.813, with each day potentially a separate offense; practitioners face professional discipline under §456.072. The statute does not define “determination” — so the clock starts at a moment nobody has instrumented.
2. The federal 60-day rule (ACA) runs in parallel for Medicare/Medicaid overpayments, and blowing it is a False Claims Act exposure — treble damages plus per-claim civil penalties.
3. All 50 states’ escheat laws then claim whatever is left unrefunded after dormancy, with a due-diligence mailing required roughly 60 days ahead of reporting for items at or above $100, and most states’ reporting deadline landing on 31 October or 1 November.

Same dollar. Three clocks. Different owners, different penalties, no shared calendar.

Meanwhile the population that has to comply is shrinking and de-staffing: independent physicians fell to 120,900 as of 1 January 2026, down 152,200 over eight years, and hospitals and corporate entities now own 63.9% of US physician practices. There are still roughly 154,787 small physician-office locations (<50 employees). These are practices with a biller, not a compliance department.

And the practitioners themselves say they are lost. On AAPC’s billing forum, sherri_g: “I am trying to find laws to help my drs understand this needs to be fixed. I am feeling lost!!” Multiple users (Jess0134, daniellecarroll7, teejae07, istanstu) show up asking for the state-specific refund rules for Florida, Illinois, Missouri and South Carolina — the uncertainty is broad, not anecdotal.

Provenance:

• Signal 1 (demand): Billing staff publicly lost on refund law and escheat. sherri_g: “I am trying to find laws to help my drs understand this needs to be fixed. I am feeling lost!!” and notatthebeach@gmail.com: “I recently was able to work our patient aging report and was able to refund tens of thousands of dollars back to patients. This report hadn’t been worked in a very long time. It was embarrassing.” — https://www.aapc.com/discuss/threads/refunding-patient-payment-credits-laws.170646/ — observed 2026-09-02
• Signal 2 (feasibility): Florida CS/CS/SB 1808 effective 1 January 2026, 30-day clock, $500/violation with each day a separate offense, “determination” undefined — https://healthcarelawmatters.foxrothschild.com/2026/01/articles/health-care-providers/floridas-new-patient-overpayment-refund-law-the-30-day-clock-for-providers-practitioners-and-facilities/ — observed 2026-09-02. Practice-management systems expose credit balances via API/report export; LLMs are now cheap enough to classify each credit’s cause (duplicate payment, COB reversal, contractual adjustment, posting error) from remittance text.
• Signal 3 (economic): Credit-balance and recovery vendors work on contingency at 15–50% depending on account age and balance size — https://www.firstcreditonline.com/healthcare-revenue-recovery/ — observed 2026-09-02. That fee structure is the opening: a contingency vendor is paid to move money, and the published risk list for credit balance files names “Over-Refunding” as a distinct compliance risk — vendors refunding money “that was a legitimate payment or an underpayment in disguise.” — https://www.jenvinhcp.com/post/5-compliance-risks-lurking-in-your-patient-credit-balance-file — observed 2026-09-02 Category: Regulatory arbitrage

3. The opportunity

Everyone in this market sells moving the money. Nobody sells the verdict on which clock owns it.

The incumbents are contingency-paid recovery and resolution vendors taking 15–50% of what they touch. That compensation structure has a built-in conflict, and the industry says so out loud: the fourth named compliance risk in a patient credit balance file is over-refunding — a vendor with a percentage incentive refunds a credit that was actually a legitimate payment, or an underpayment in disguise, and the practice loses real revenue. A vendor paid on volume moved cannot credibly sell you the answer “don’t refund this one.”

The other half of the market is reporting software. ReportMyUP (Sovos) starts at $299/year and files NAUPA-compliant reports across all 50+ states; HRS Pro does the same job. These are good products and I am not going to beat them at filing. But they file what you hand them. They do not tell you which of your 900 credits is a Medicare overpayment on a 60-day federal fuse, which is a Florida patient overpayment on a 30-day state fuse, which is a $4 credit you should apply to the patient’s next copay, and which is a posting error that should be reversed rather than refunded at all.

That triage is the product. It is a classification problem sitting between two well-served adjacent markets, and both adjacent players are structurally disqualified from doing it — the contingency vendor by its incentive, the filing software by its position at the end of the pipe.

The current alternative is a biller opening the credit-balance report once a month and eyeballing it. The forum quote about refunding “tens of thousands of dollars” from an aging report that “hadn’t been worked in a very long time” is what that alternative produces.

4. Target market

• Primary customer: Practice manager or lead biller at an independent physician, dental or therapy practice in the US — 1–10 providers, <50 employees, running Athenahealth, eClinicalWorks, Kareo/Tebra, DrChrono or AdvancedMD. Beachhead: Florida, because SB 1808 gave Florida practices a dated, penalised 30-day clock the rest of the country doesn’t have yet.
• Why they buy: In their words — “I am feeling lost!!” and “It was embarrassing.” They know the credit report is unworked, they know there’s a law, they don’t know which law applies to which line, and they’re one complaint or one audit away from finding out. notatthebeach@gmail.com again: “If nothing else, it’s unethical to keep someone’s money if they overpaid. Especially if it’s in the hundreds of dollars.” The pain is felt monthly, at close.
• Rough TAM reasoning: ~154,787 small physician-office locations (<50 employees) in the US, plus dental and behavioural-health practices carrying the same duty. Serviceable beachhead: Florida practices in that band, plausibly 8–12K locations, expanding state by state as more states copy the 30-day pattern.
• Why now for them: Florida’s clock started 1 January 2026 and is enforced per-day. The federal 60-day rule and escheat obligations were always there but had no forcing function for a small practice; a dated state statute with per-violation fines is the forcing function.

5. Product sketch (MVP)

• Pull the credit-balance / patient-aging report from the practice’s PM system (export first, API where available) and normalise it into one list.
• Classify each credit by cause — duplicate payment, coordination-of-benefits reversal, contractual adjustment, upfront collection over-estimate, posting error — from the remittance and adjustment text.
• Assign each credit to the clock that owns it: federal 60-day (government payer overpayment), state refund statute (e.g. Florida 30-day), escheat dormancy, or “not a refund — reverse the posting.”
• A dated worklist ordered by fuse length, showing days remaining on each item and what happens when it burns.
• Hold recommendations with reasons: patient has an appointment inside 90 days and the credit is under the practice’s threshold; credit is an underpayment in disguise; charge correction pending.
• Generate the due-diligence letter for escheat-bound items at or above the state’s minimum, timed to land inside the required window before the reporting deadline.
• Export a clean NAUPA-ready file of genuinely escheatable items — to hand to ReportMyUP/HRS Pro, not to compete with them.
• A per-credit decision log with timestamps: what we determined, when we determined it, why. This is the “determination” record Florida’s statute leaves undefined.

6. AI angle — what’s load-bearing

The classification is the product, and it is genuinely a language problem. A credit balance does not arrive labelled. It arrives as a negative balance with a trail of remittance advice, adjustment codes, payer names and free-text posting notes that differ per PM system, per payer and per biller’s habits. Deciding whether a −$212.40 is a duplicate commercial payment, a Medicare overpayment, a COB recovery, or a biller fat-fingering an adjustment requires reading that trail the way an experienced biller reads it.

Remove the AI and this is a rules engine that demands the practice already know each credit’s cause — which is exactly the work they can’t do. The rules layer (which statute, how many days, what dormancy) is deterministic and should be hand-coded and auditable; the AI does the reading and the causal call, and hands the rules engine a typed input. That split is also the honest one for a compliance product: never let the model decide the deadline, only what kind of thing it’s looking at.

7. Localization angle (if any)

N/A — this is a US play, and deliberately a state-by-state one. The localization that matters here is jurisdictional, not linguistic: Florida’s 30-day statute, California’s 30-day-on-request rule, and 50 separate escheat regimes with different dormancy periods, due-diligence minimums and deadlines. That per-state rule library is the thing that takes work to assemble and is the closest thing this product has to a moat. Spanish-language patient letters are worth shipping for Florida and Texas, but that’s a feature, not a wedge.

8. Business model — path to $1M–$5M ARR

• Pricing: $149/mo for 1–3 providers, $299/mo for 4–10, $599/mo for multi-location. Flat subscription, explicitly not contingency — “we don’t take a cut of your refunds” is the sales line, because the incumbent’s cut is the thing that makes them untrustworthy on the hold decision.
• ACV: ~$2,700 blended.
• Rough math to $1M ARR: 370 practices at $2,700 = $1.0M. That’s ~4% of a plausible 8–12K Florida beachhead, before leaving the state.
• Rough math to $5M ARR: ~1,850 practices, which needs the state rule library covering the top 10–15 states and a channel through billing companies rather than one practice at a time. A billing company running 40 practices is one sale and 40 seats.
• Expansion path: more locations, then dental and behavioural health on the same rules engine, then an annual escheat-filing add-on. The natural upsell is the reverse direction — practices that start with the monthly triage will ask you to just handle the November filing too.

9. Go-to-market wedge — first 100 customers

• Florida first, dated and specific. Florida licenses practitioners by board and publishes licensee data; medical and dental practice lists are also purchasable from standard healthcare data vendors. Build a list of independent FL practices in the 1–10 provider band, email the practice manager with the actual statute cite and the per-day fine, and offer a free read of their credit report. The email writes itself because the law is real, dated and named.
• AAPC forum and the billing-credential community. The threads cited above are people asking, in public, for exactly this answer, state by state. Answer the state-law questions properly and for free — publish a per-state refund-and-escheat sheet — and the tool is the natural next click. sherri_g and everyone asking about FL/IL/MO/SC are literally raising their hands.
• Billing companies as the channel. Outsourced medical billing runs at 4–9% of net collections; a billing company carrying 30–60 small practices has this exposure multiplied and no good answer for it. One signed billing company is 30–60 practices. This is the path to $5M and it should start in month 2, not month 12.
• The free credit-report audit as the demo. Ask for a CSV export, return a categorised list with the fuse on each item and a total dollar figure sitting in the wrong bucket. The “tens of thousands of dollars” quote suggests the first read on an unworked report is frequently alarming, and alarm closes.
• State-law content that ages well. One page per state on the refund clock, the escheat dormancy, the due-diligence minimum and the deadline. This is the query these people are already typing.

10. Build complexity — justification

Medium. The PM-system integrations are the real work — Athenahealth, eClinicalWorks and Tebra all differ, and CSV export is the honest v1 for most of them rather than fighting for API access up front. The rules library is 50 states of statutory detail that has to be assembled by hand and kept current; that’s tedious but bounded, and it’s also the moat. The AI classification sits on off-the-shelf models. A pair could ship a credible Florida-only v1 on CSV import in 10–12 weeks, then add states and direct integrations.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Helping a provider meet refund and escheat duties. No licensed activity — the product recommends, the practice decides.
Ethical — no harm / dark patterns	✅	The product’s bias is toward returning patients’ money on time. The “hold” recommendation needs care: it must never be a tool for sitting on refunds, and holds should expire loudly.
Market exists (evidence above)	✅	Contingency vendors at 15–50%, filing software at $299/yr, and billers publicly asking for the answer.
1–5 person team can build this	✅	Pair, 10–12 weeks to a Florida v1.
Launchable with <$50K / ₹40L	✅	Data list, inference costs, and a healthcare attorney’s review of the state rule sheets.

All five pass.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	16/20	Dated statute, per-day fines, FCA tail on the federal side. Felt monthly at close. Docked because most small practices have absorbed the risk quietly for years — the pain is real but has been survivable, and Florida is currently the only hard forcing function.
Demand evidence	15	12/15	Verbatim billers asking for exactly this, priced incumbents on both flanks, sized population. Not 13+ because I found no one paying specifically for triage today — I’m inferring willingness to pay from adjacent spend.
Build feasibility	15	11/15	CSV-first v1 is genuinely simple; PM integrations and 50 states of rules are the grind.
Distribution clarity	15	12/15	Named state, named statute, named forum, and billing companies as a multiplier. Docked because practice managers are famously hard to reach cold.
Revenue mechanics	15	11/15	$149–599/mo is well inside what practices pay for RCM tooling, and flat-fee vs contingency is a clean pitch. $1M needs 370 practices, which is real work.
Time to first revenue	10	8/10	The free credit-report audit converts fast when the number is ugly; expect first paid inside 6–8 weeks of launch.
Defensibility	10	6/10	The 50-state rule library plus accumulated per-credit decision logs create switching cost by month 12. But a determined RCM incumbent could build this; the protection is that the contingency players don’t want to and the filing players sit downstream.
Total	100	76/100

13. Qualitative modifiers

Founder-fit tags

domain-expertise-required · sales-heavy

You need someone who has actually worked a credit-balance report, or a billing-company partner who has. Getting the hold/refund/reverse call wrong in either direction is how this product loses trust — over-refund and you cost them revenue, under-refund and you’ve put them on the wrong side of a statute.

Key assumptions to validate (3–5)

1. Assumption: A small practice’s credit-balance report is materially misclassified — enough items in the wrong bucket that the first read is alarming. How to test: Get 10 practices to hand over an anonymised credit export; classify by hand; measure what fraction is on a clock the practice didn’t know about.
2. Assumption: Practices will pay a flat $149–299/mo for a verdict rather than give a contingency vendor 15–50% of the movement. How to test: Put both offers in front of 30 Florida practice managers and see which they pick, and whether the anti-contingency framing actually lands.
3. Assumption: Florida’s undefined “determination” trigger makes practices want a timestamped decision log. How to test: Ask 20 FL practices what they currently record when they identify an overpayment. If the answer is “nothing” and they shrug, the audit-trail value prop is weaker than I think.
4. Assumption: Billing companies will resell or embed this. How to test: Pitch 10 billing companies in months 2–3; one signed pilot validates the $5M path.

Risk flags

1. Regulatory-drift risk: 50 states, each free to change dormancy periods, minimums and deadlines. The rule library is the moat and also a permanent maintenance tax. If it goes stale the product is worse than useless — it’s confidently wrong about a deadline.
2. Incumbent expansion: Sovos sits directly upstream of this with an existing SMB product and channel. If ReportMyUP adds classification, the wedge narrows fast. The counter is depth on the clock-triage and the decision log, not breadth.
3. Liability posture: Telling a practice “don’t refund this one” is a recommendation with consequences. This needs clear framing as decision support, an attorney-reviewed rules layer, and E&O cover. Do not let this drift into practising law or accounting.
4. Market timing / consolidation: The customer base is actively shrinking — 152,200 fewer independent physicians over eight years, 63.9% of practices now corporate-owned. Every acquired practice leaves the segment. This is a market with a clock on it, which argues for moving fast and for the billing-company channel over one-at-a-time sales.

14. Structured verdict

Score:                  76/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical founder paired with an RCM/billing domain expert
Time to revenue:        6–8 weeks post-launch
Capital to launch:      $15–25K (data lists, inference, attorney review of state rule sheets)
Top 3 assumptions to validate first:
  1. Credit reports are materially misclassified — hand-classify 10 real practice exports
  2. Flat fee beats contingency in the practice manager's head — offer both to 30 FL practices
  3. Billing companies will carry it — pitch 10, need 1 pilot
Kill criteria:
  - Abandon if <3 of 10 hand-classified credit reports show material misclassification
  - Abandon if Sovos/ReportMyUP ships credit classification before the Florida v1 lands
  - Abandon if <5% of 200 targeted Florida practice managers accept a free credit-report audit

15. Next step — 1-week validation sprint

• Day 1–2: Write the Florida one-pager — SB 1808, the 30-day clock, the $500/day exposure, the federal 60-day overlay, the escheat tail. Build a list of 200 independent Florida practices in the 1–10 provider band.
• Day 3–4: Email all 200 offering a free credit-balance read. Simultaneously post genuinely useful state-law answers in the AAPC threads where people are already asking for FL/IL/MO/SC.
• Day 5: Hand-classify every export received. Measure two numbers: how many practices accepted the free audit, and what fraction of credits in those exports sat on a clock the practice hadn’t identified.

Falsifiable outcome: ≥10 of 200 accept the audit (5%), and ≥30% of credits in the exports received are misclassified or unassigned to any clock. Below either threshold, the triage isn’t a felt problem and I drop it — the practices are surviving on the monthly eyeball, and a statute they’ve never been fined under won’t change that.