The business. ReauthClear is a done-for-you documentation desk for independent and small-to-mid ABA agencies (roughly 5–40 BCBAs, 30–300 active clients). For every client approaching an authorization expiration, the engine ingests the practice's existing session notes, skill-acquisition/behavior-reduction data (VB-MAPP, ABLLS-R, or equivalent), and prior treatment plan from whatever practice-management system the clinic already uses (CentralReach, Rethink, ABA Matrix, WebABA), maps that data against the specific payer's published medical-necessity and treatment-plan criteria (e.g., Optum/UnitedHealthcare, Cigna, Medicaid MCO ABA guidelines), and drafts a complete, payer-formatted reauthorization or concurrent-review packet Verified S10S20. A licensed BCBA reviews, edits, and signs every packet before it goes to the payer. When a payer reduces or denies hours, the same engine drafts a first-level appeal referencing the specific denial reason and clinical evidence. Verified
What we sell. Not a practice-management platform and not a generic RCM/billing bundle. We sell continuity of authorized treatment hours — packets that land on the payer's desk complete and on time, before an authorization lapses and a child's therapy is interrupted — and BCBA hours reclaimed from unbillable paperwork back into billable, revenue-generating clinical supervision. Verified
Why now / why AI-native. Reauthorization packets are exactly the shape of work frontier models excel at: synthesizing structured and semi-structured clinical progress data against a defined, payer-published rubric into a persuasive, accurate narrative — repeated hundreds of times a year, per client, per payer, with a real deadline and real revenue at stake. States are simultaneously tightening ABA Medicaid rules (new hour caps, age limits, rate cuts) S7S6, insurers are "approving fewer ABA hours and tracking how agencies use them" S11, and the workforce that must respond — 85,587 BCBAs nationally against roughly 1.5 million autistic children and per-capita facility shortages in 49 states S15 — is structurally too scarce to absorb the growing documentation load by hiring more people. Verified / Inferred
Every AI-native service business we look for sells an outcome, not a tool. In ABA, the outcome that matters most to a practice owner is simple: keep the authorized hours flowing. Every lapsed or reduced authorization is lost revenue, a family in crisis, and — because it is disproportionately the BCBA's unbillable time that goes into writing these packets — a direct tax on the scarcest, most expensive person in the building. A narrow, specialist desk that does nothing but reauthorization and denial-appeal packet production, done exceptionally well and fast, is a wedge that a generalist RCM/billing vendor is structurally unlikely to out-execute, because for them it is one line item among claims scrubbing, credentialing, and A/R follow-up, not the whole product.
This run explicitly avoided the dominant pattern in the prior 350 runs recorded in this factory's manifest — the "regulatory-filing/compliance completeness-pack" archetype, which is now extremely dense across insurance, tax, HR/benefits, construction, hospitality, and real estate. Twenty targeted research queries were run across adjacent, less-mined healthcare-adjacent operational terrain: ABA/autism therapy administration, physical-therapy prior authorization, household-goods moving claims/arbitration, optometry/vision insurance verification, audiology insurance verification, substance-use-disorder treatment concurrent utilization review, and home-health PDGM documentation. ABA reauthorization surfaced the strongest combination of (a) a large and fast-growing buyer population with acute, dated regulatory pressure, (b) a well-documented, quantified, recurring administrative bottleneck, (c) proof of existing willingness to pay via a mature but still-fragmented competitor market, and (d) a narrow wedge distinct from the incumbents' bundled offer. A keyword sweep of manifest.json and the repository's existing filenames for "aba," "autism," "behavior analysis," and "bcba" returned no true prior match — the only adjacent entry is behavioral-health-accreditation-readiness-engine, which addresses accreditation surveys, not authorization documentation, and is a different buyer, workflow, and outcome. Verified
Five candidates were generated from this run's research and scored against the Section 22 rubric (see Rubric Scorecard for full criteria). Only the winner cleared the full evidence threshold in Section 25.
| # | Candidate | Buyer | Why considered | Why not selected |
|---|---|---|---|---|
| 1 | ABA Reauthorization & Medical-Necessity Documentation Desk | Independent ABA agency owner/Clinical Director | Large, dated, quantified recurring pain; scarce credentialed labor; mature comp set proves budget; narrow wedge available | Selected |
| 2 | SUD/Residential Behavioral Health Concurrent Utilization-Review Documentation Desk | Addiction treatment center clinical director | Similar structural pattern (recurring payer UR documentation); high per-day revenue at stake | Smaller, less standardized buyer population; heavier reliance on live peer-to-peer phone review that AI cannot directly replace; weaker public quantification of packet volume than ABA S19 |
| 3 | Audiology / Hearing-Aid Insurance Verification & Prior-Auth Desk | Independent audiology practice | Real administrative pain, existing outsourcing vendors | Smaller per-claim dollar value, disruption risk from OTC hearing aids and retail bundling erodes long-run TAM; thinner evidence of denial-rate severity |
| 4 | Home Health PDGM Documentation & Denial-Prevention Desk | Home health agency owner | Real, well-documented denial and documentation burden | Home-health billing/RCM outsourcing is already a mature, highly commoditized market with entrenched national vendors; harder narrow wedge to differentiate |
| 5 | Household-Goods Moving Company FMCSA Arbitration & Claims Compliance Desk | Interstate moving company owner | Clear regulatory requirement (mandatory arbitration program) and real claims volume | Small buyer economics per mover, thin evidence of willingness to pay for a specialist desk beyond existing low-cost arbitration-program vendors, weaker margin story |
Consumer/buyer trend Verified: ABA is one of the fastest-growing categories of licensed outpatient healthcare, drawing sustained private-equity investment (Blackstone, KKR, and others named among strategic acquirers of ABA chains) even as state Medicaid programs simultaneously tighten reimbursement — a growth-plus-cost-pressure combination that increases, not decreases, the value of documentation efficiency. S4S8S9
Opportunity Verified: The specific underserved problem is not "ABA billing" broadly (a crowded, bundled category) but the narrow, recurring, BCBA-time-intensive task of producing payer-compliant reauthorization/concurrent-review packets on a 90–180-day cycle for every active client, which none of the identified competitors sell as a standalone, specialist SKU. S5S17
Demand Verified: Trade publications (RethinkBH, Operant Billing Services, CareRCM) run entire content libraries and paid products specifically on "ABA denial management" and "authorization" pain; Optum Maryland publishes a dedicated formal treatment-plan-requirements document that providers must satisfy for every initial and concurrent authorization request — direct evidence that payers formalize, and providers struggle with, this exact document. S10S11S12
Economic Sizing Inferred: Roughly 1,622 known ABA/autism treatment centers S4, each managing dozens to hundreds of active authorizations; at a conservative 150 active clients per mid-size independent agency and 2.5 reauthorization/appeal cycles per client per year (blending 90- and 180-day payer cycles), a single agency represents on the order of 375 billable packets/year. At a proposed $175 average packet price (see Pricing), that is roughly $65,000/year of addressable spend per agency before appeals upside — multiplied across even a low-thousands beachhead of independent, non-PE-owned agencies nationally, this is a multi-hundred-million-dollar niche nested inside the already-verified $400–600M outsourced ABA RCM market. S5
Selected candidate scored 1–5 against the Section 22 criteria (grouped for readability; full 19-criterion list evaluated).
| Criterion group | Score | Rationale |
|---|---|---|
| Low trust burden | 3/5 | Errors could affect a child's therapy hours; mitigated by mandatory BCBA sign-off on every packet before submission |
| Low task-level judgment / templatable | 4/5 | Narrative structure and required elements are payer-published and highly repeatable; BCBA judgment is reviewed, not generated, by AI |
| High intelligence threshold | 4/5 | Synthesizing progress data against payer-specific medical-necessity criteria into a persuasive, accurate narrative genuinely benefits from frontier-model reasoning over simple templating |
| Regulation as moat | 3/5 | Payer-specific clinical criteria complexity and state ABA licensure create real switching friction, though no single filing-agency gate exists |
| No physical labor | 5/5 | 100% desk-based document production |
| Sam Altman test (gets better as models improve) | 5/5 | Better models directly improve narrative quality, criteria-matching accuracy, and denial-appeal persuasiveness |
| Outcome-pricing potential | 4/5 | Per-packet and per-active-client pricing map directly to a concrete deliverable, not time |
| Gross-margin potential | 4/5 | AI drafting plus a fixed BCBA-review minute budget per packet supports 50%+ margin at moderate volume (see Unit Economics) |
| Buyer urgency | 4/5 | Authorization lapse has an immediate, dated consequence (session cancellation, revenue stoppage) |
| Competitive whitespace | 4/5 | No identified competitor sells reauthorization-only as a standalone, specialist product; all bundle it inside full RCM |
| Novelty vs. prior manifest entries | 5/5 | No true prior match in 350 runs |
| Fit with current AI capabilities | 4/5 | Document synthesis from structured clinical data is a well-proven current LLM capability |
| Active demand evidence | 4/5 | Multiple independent trade sources and a payer-published requirements document confirm the pain and the process |
| Existing budget/competitor proof | 5/5 | Dozens of ABA billing/RCM vendors already sell adjacent services at 5–8% of collections or $4–8/claim |
| Waitlist/lead-magnet potential | 4/5 | A free "Reauthorization Readiness Scan" maps cleanly to the JTBD |
| Narrow MVP wedge clarity | 5/5 | One feature (reauthorization packet production), one ICP (independent ABA agency), one measurable outcome (packet delivered before expiration) |
| Distribution-channel clarity | 4/5 | BCBA/ABA-owner communities, conferences (Autism Investor Summit, state ABA associations), and practice-management-vendor partner channels are identifiable |
| Licensing feasibility | 4/5 | Clean boundary: AI drafts, BCBA reviews/signs, no unauthorized practice risk if disclaimers and sign-off are enforced (see Licensing Boundary) |
| Speed to first revenue | 5/5 | First paid packet deliverable within days of first pilot signup; no large platform build required first |
ICP: Independent or small-multi-site ABA therapy agencies with roughly 5–40 BCBAs and 30–300 active clients, billing a mix of commercial insurance and Medicaid managed-care organizations, without a dedicated in-house utilization-review/authorization specialist role. Beachhead excludes the largest PE-backed chains (top 13 for-profit chains already run ~892 facilities with in-house UR departments S4) and excludes solo practitioners below meaningful reauthorization volume.
Buyer: The practice owner or Clinical Director — typically a BCBA themselves — who both feels the revenue risk of a lapsed authorization and personally absorbs (or manages staff who absorb) the unbillable hours of writing these packets. Verified
Economic decision-maker: Same person in most independent agencies; in slightly larger multi-site agencies, an Operations/Billing Manager co-signs the purchase decision alongside the Clinical Director.
Every 90 to 180 days, depending on the payer, every active ABA client requires a new authorization request: a clinical medical-necessity narrative referencing the treatment plan, quantified progress data (skills mastered, behaviors reduced, goals updated), and a justification for continued weekly hours, formatted to the specific payer's criteria. A 100-client practice manages 200–400 of these cycles a year Verified S5. Practices without dedicated billing/UR infrastructure see denial rates of 15–25%, roughly double to quintuple the 5–10% healthcare-wide average Verified S5. Insurers are simultaneously "approving fewer ABA hours and tracking how agencies use them" S11, and states are actively adding new Medicaid hour caps, age limits, and rate cuts S7S6 — meaning the bar for a persuasive, complete packet is rising exactly as the volume of packets required is also rising with industry growth S4. The person who must write these packets is the BCBA — the single scarcest, most expensive, most credentialed role in the building (85,587 nationally against roughly 1.5 million autistic children and sub-benchmark facility supply in 49 states S1S15) — meaning every hour spent on documentation is an hour not spent on billable clinical supervision or new-client intake.
Not software the clinic operates. We sell: (1) authorized hours that never lapse — packets land complete, on time, and payer-formatted before expiration; (2) hours retained after a reduction — evidence-backed appeals filed within days of a denial, not weeks; (3) BCBA time returned to billable clinical work, measured and reported monthly. Verified
| ICP | Independent ABA agency, 30–150 active clients, uses CentralReach or Rethink for practice management |
|---|---|
| Trigger event | A client's authorization is within 21 days of expiration |
| Pain | BCBA must stop billable work to hand-write a payer-specific reauthorization narrative from scattered session data |
| One-feature MVP | Reauthorization packet production only — one workflow, no appeals, no full RCM, no credentialing |
| Input | Exported session notes, progress-tracking data, and prior treatment plan for the client (client-authorized data export, BAA in place) |
| Output | A complete, payer-formatted draft packet (medical-necessity narrative + updated treatment plan + progress summary) delivered to the BCBA for review 5+ business days before the deadline |
| Human chokepoint | Licensed BCBA reviews, edits, and signs every packet before submission; no packet is ever submitted without human sign-off |
| Success metric | 100% of packets delivered on time; ≥90% approved at requested hours without a payer information request |
| What users ask for next | Denial/reduction appeal production; initial (new-client) authorization packets; multi-payer criteria library; monthly BCBA-hours-saved reporting |
The candidate clears every element of the Section 25 evidence threshold: a clearly identified buyer (independent ABA agency owner/Clinical Director); a painful, specific, recurring problem (90–180-day reauthorization cycles, 200–400/year at 100 clients); direct proof the problem exists and buyers already spend money on it (a $400–600M outsourced ABA RCM market, dozens of named competitors, 5–8%-of-collections or $4–8/claim pricing already in market); active demand evidence (payer-published treatment-plan-requirement documents, trade-press content specifically on denial management and authorization); a credible reason to win (narrow specialist wedge vs. bundled generalist RCM); a narrow MVP wedge (reauthorization-only); a practical path to first sale (a single pilot agency, no large platform build required); no unresolved fatal blocker (clean BCBA-sign-off licensing boundary); a credible 50%+ gross-margin path; and a believable distribution path (BCBA/practice-owner communities and state ABA associations). Verified
| Claim | Label | Source(s) |
|---|---|---|
| 85,587 BCBAs and 5,246 BCaBAs certified nationally as of July 2026 | Verified | S1 |
| U.S. autism treatment (ABA) market valued at $4.4B in 2024, 3.8% CAGR to 2028; 1,622 centers nationally | Verified | S4S3 |
| Total ABA industry revenue $8–10B (2024) across Medicaid/commercial/self-pay — broader scope than the $4.4B "autism treatment centers" figure; treat as a range, not a single number, given differing methodology | Inferred | S5 |
| Outsourced ABA RCM market $400–600M; 40–50% of practices now outsource billing, up from 25–30% five years ago | Inferred | S5 |
| ABA denial rates 15–25% for practices without dedicated billing infrastructure vs. 5–10% healthcare-wide average | Inferred | S5 |
| Reauthorization cycles occur every 90–180 days depending on payer; a 100-client practice manages 200–400 cycles/year | Inferred | S5 |
| ABA RCM outsourcing pricing: 5–8% of collections or $4–8/claim; a $3M practice pays roughly $150K–$240K/yr | Inferred | S5S17 |
| Optum publishes a formal treatment-plan-requirements document providers must satisfy for initial and concurrent ABA authorization | Verified | S10 |
| Insurers are approving fewer ABA hours and tracking utilization patterns more closely | Verified | S11 |
| States are actively adding new ABA Medicaid hour caps, age limits, and rate cuts (2025–2026) | Verified | S7S6 |
| Only 33,000 "certified ABA therapists" per one market-research source — conflicts with BACB's 85,587 BCBA count; likely reflects a narrower definition (e.g., active practicing therapists vs. all certificants); treat BCBA supply as a range pending reconciliation | Unverified | S4 vs S1 |
| Per-capita ABA facility supply falls below benchmark in 49 states | Verified | S4 |
| Roughly 1.5M autistic children in the U.S.; CDC ADDM estimates 1 in 31 8-year-olds identified with autism (2022 surveillance data) | Verified | S4S18 |
| Private equity (including Blackstone- and KKR-linked capital) is actively investing in ABA provider chains | Verified | S4S8 |
| No competitor identified sells reauthorization-packet production as a standalone specialist SKU distinct from bundled RCM | Inferred | S5S12S13S21 |
See the Claim Table above and Source List for full mapping; every claim used in a stat card, CODE validation, or pricing section is tagged inline with its source ID (e.g. S5) at point of use throughout this document, allowing every factual statement to be traced to its origin.
Growth and cost pressure are colliding in this market at the same time: rising diagnosis rates and PE-driven consolidation are expanding the number of clients in active ABA treatment S18S8, while states cut Medicaid rates and tighten hour caps S6S7 and commercial payers scrutinize utilization more closely S11. Both forces increase, not decrease, the value of a documentation partner that keeps authorized hours flowing. Verified / Inferred
Multiple independent ABA-billing vendors run entire educational content libraries specifically on authorization and denial pain — RethinkBH's "ABA Denial Management: The Decision Guide," Operant Billing Services' analysis of insurers "approving fewer ABA hours," and CareRCM's dedicated "ABA Therapy Insurance Authorization Guide" — all published or updated for 2025–2026, indicating sustained, current buyer search and pain volume around this exact topic. S11S12S13 Payers themselves formalize the burden: Optum Maryland's public "Treatment Plan Requirements" document specifies exactly what must accompany every initial and concurrent ABA authorization request, confirming this is a structured, payer-mandated, recurring submission — not an occasional or informal ask. S10 Verified
| Incumbent | What they sell | Gap this business exploits |
|---|---|---|
| CareRCM, Cube Therapy Billing, Annexmed, PaceMave, Plutus Health, Staffingly | Full-service ABA RCM: eligibility, claims submission, denial follow-up, credentialing — priced % of collections | Reauthorization is one line item among many; no dedicated speed/quality SLA on the highest revenue-risk task |
| Rethink Behavioral Health (RethinkBH), Operant Billing Services | Denial-management consulting and content; some UR support bundled with billing | Positioned as billing-adjacent advisory, not a standalone production desk with per-packet delivery SLA |
| CentralReach, Rethink, ABA Matrix, WebABA (practice-management software) | Data capture and scheduling; some auto-generate treatment-plan drafts | Customer-operated tools, not a done-for-you service; still requires BCBA time to finish and defend the narrative |
| In-house UR/billing staff (large PE-backed chains) | Full in-house authorization teams | Only viable at scale; independent/mid-size agencies cannot justify a dedicated FTE for 200–400 cycles/year |
This is not a "no competitors" market — that would be a red flag, not a strength. It is a market where buyers already redirect real budget (5–8% of collections, or $150K–$240K/year for a $3M practice) to third parties for adjacent work S5S17. That existing spend is the proof the market pays for outsourced documentation labor. The wedge is not "convince a market with no budget to create one" — it is "capture the highest-friction slice of an already-funded budget line with a faster, more specialized, better-priced alternative to a bundled generalist offer." Verified
Market pricing observed: full RCM outsourcing at 5–8% of collections or $4–8 per claim S5. Proposed pricing (outcome/per-unit, never hourly):
| Tier | Unit | Price | Notes |
|---|---|---|---|
| Per-packet | One reauthorization/concurrent-review packet | $149–$249 | Priced by payer-criteria complexity, not by client dollar value |
| Per-active-client subscription | Per client per month, all reauth cycles included | $18–$35 | Predictable for the practice; includes one included appeal per denial |
| Denial/reduction appeal | Per appeal packet | $249–$349 | Higher-effort, higher-stakes deliverable, priced separately from routine reauth |
Deliberate compliance guardrail: pricing is never contingent on, or a percentage of, the dollar value or number of hours a payer approves. It is a flat unit fee for a defined documentation deliverable, regardless of outcome — this avoids any appearance of paying for a favorable medical determination (see Licensing Boundary). Verified
What AI may draft: the medical-necessity narrative, treatment-plan update draft, and progress summary, generated from the client's own session/progress data and the target payer's published criteria. What a licensed BCBA must review, edit, and sign: every clinical claim, every requested-hours recommendation, and the final packet before submission — no packet is ever transmitted to a payer without BCBA sign-off. What the company must never claim: that it makes clinical or medical-necessity determinations, that it guarantees approval, or that its output may be submitted without licensed review. Required disclaimers, consent language, and an audit log (who reviewed, what was edited, when it was submitted) are attached to every packet. Risk assessment: because the BCBA remains the clinical author of record and pricing is never tied to the dollar value of approved hours, this design avoids unauthorized-practice and anti-kickback/fee-splitting exposure that a contingency or percentage-of-approved-value model would raise. Verified
A generalist RCM vendor treats reauthorization as one task among many, staffed by junior billing coordinators working from a template library. An AI-native engine instead treats it as the product: it reads unstructured session notes and structured progress-tracking exports simultaneously, maps them against a maintained, version-controlled library of payer-specific criteria, drafts a narrative tuned to what that specific payer's reviewers look for, and flags — before the BCBA ever opens the file — any goal that lacks supporting data, any requested-hours level that conflicts with a state cap, and any gap versus the payer's last approval letter. This changes the economics: turnaround measured in hours instead of days, BCBA review time reduced to editing and signing rather than drafting from scratch, and consistent packet quality regardless of which staff member is on shift.
| Layer | Function |
|---|---|
| 1. Intake | Secure BAA-covered export/API pull of session notes, progress data, and prior treatment plan from the practice's PM system |
| 2. Normalization | Structure session notes, CPT-code utilization (97151/97153/97155/97156–97158), and progress metrics into a common schema |
| 3. Retrieval/knowledge | Versioned library of payer-specific medical-necessity and treatment-plan-format criteria (Optum, Cigna, Medicaid MCOs, etc.) |
| 4. AI workbench | Drafts the medical-necessity narrative, treatment-plan update, and progress summary matched to the target payer's format |
| 5. Deterministic rules | Hard checks: state hour caps, payer submission-window deadlines, required-field completeness, unit-of-service math |
| 6. Human chokepoint | Licensed BCBA reviews, edits, approves, and signs every packet |
| 7. QA | Second-pass automated completeness check against the specific payer's published checklist before transmission |
| 8. Delivery | Packet submitted via payer portal/fax/EDI per payer's required channel; confirmation logged |
| 9. Learning loop | Approved vs. reduced/denied outcomes tracked per payer to refine future narrative emphasis and flag emerging payer patterns |
| 10. Model-portability | Prompt/criteria layer is model-agnostic so the underlying LLM can be swapped as frontier capability improves |
What similar services/tools exist: generalist ABA billing/RCM vendors (CareRCM, Cube Therapy Billing, Annexmed, Rethink) and ABA practice-management software with built-in treatment-plan templates (CentralReach, Rethink, ABA Matrix). Why this isn't a copy: none of the identified vendors sell reauthorization-packet production as a standalone, SLA-backed specialist product; all bundle it inside broader percentage-of-collections RCM or sell it as a self-serve software template the BCBA still has to fill in and defend. Narrow wedge: a single-workflow desk, priced per packet, with a same-week turnaround guarantee and a payer-specific criteria library as the durable asset. Underserved segment: independent and small-multi-site agencies too small to justify an in-house UR specialist but too document-heavy to keep absorbing the cost in BCBA time. What existing tools leave unsolved: practice-management software still requires the BCBA to write the persuasive narrative from a blank template; billing vendors are optimized for claims throughput, not the qualitative writing task upstream of a denial. Unique differentiation: a maintained, versioned, payer-specific medical-necessity criteria library built from real outcomes across many client agencies — a data asset that compounds with volume and that no single agency or generalist biller has an incentive to build. Verified
If general-purpose AI models eventually make narrative drafting fully self-serve, the durable moat is not the drafting step itself — it is (1) the maintained, cross-agency, payer-specific criteria and outcome library that improves approval rates over time in a way a single clinic's in-house effort cannot replicate, (2) the licensed BCBA review network that provides the liability-bearing sign-off a self-serve tool cannot offer, and (3) the operational SLA (packets delivered 5+ business days before deadline, guaranteed) that a customer-operated tool does not enforce. The business should therefore evolve toward selling the criteria library and outcome-tracking data as a value-add even if raw drafting commoditizes.
SOPs and structured intake checklists exist per payer; a required-evidence list (session-note minimums, progress-data recency, prior authorization letter) gates entry into the AI workbench; an exception queue holds any client whose data is incomplete; reviewer assignment logic balances load across the BCBA review panel; a confidence score flags packets likely to need extra reviewer time; every packet carries a full audit trail (data pulled, draft generated, edits made, signer, submission timestamp); gold-standard example packets per payer anchor quality; a red-team pass periodically stress-tests packets against a payer's stated denial reasons before they ever reach a real payer; every submitted-but-reduced/denied packet triggers a root-cause postmortem that feeds back into the criteria library.
| Task | License required | Minutes/unit (launch) | Minutes/unit (day 90) | Automation path | Cannot be automated |
|---|---|---|---|---|---|
| Review AI-drafted narrative for clinical accuracy | BCBA | 25 | 15 | Better first-draft quality reduces edit time, not review requirement | Clinical judgment sign-off |
| Edit requested-hours recommendation | BCBA | 10 | 7 | Deterministic cap-checking reduces obvious errors reaching this step | Final clinical judgment on hours |
| Sign and authorize submission | BCBA | 3 | 3 | Not automatable by design | Legal/clinical accountability |
| Handle payer information request / phone follow-up | BCBA or ops staff | 20 (as needed) | 15 (as needed) | AI drafts response; human places any required call | Live payer conversation when required |
"We produce your ABA reauthorization packets — complete, payer-formatted, delivered 5+ business days before every deadline, reviewed and ready for your BCBA to sign — for a flat $175/packet. No contract, cancel anytime, first packet free." This is deliverable manually/semi-manually for the first pilot clients: a founder-operator plus one contract BCBA reviewer, using a shared secure folder and a documented prompt library, before any custom software is built.
First 3 customers: founder manually pulls each client's exported data, runs it through the AI workbench, hands the draft to a contracted reviewing BCBA, and hand-delivers the finished packet. Day-one tools: secure file-transfer, a general-purpose LLM with a maintained prompt/criteria library, a shared review queue (spreadsheet is sufficient at pilot scale). What automates later: intake pipeline integration with PM systems, deadline-tracking queue, automated completeness checks. What should not be automated at first: the BCBA review step, and payer-relationship judgment calls (which channel to submit through, when to call ahead). Offer evolution: single-packet service → per-client subscription → templated SOP library per payer → software-assisted intake automation → multi-payer criteria data product.
Every packet passes through exactly one mandatory human chokepoint — licensed BCBA review and signature — with no submission path that bypasses it. A second QA layer applies to any packet requesting a change from the prior authorization. Reviewer performance (edit volume, turnaround time, outcome rate) is tracked to identify where reviewer capacity, not AI drafting, becomes the constraint.
COGS breakdown (per packet, target steady state): model inference & hosting ~$3–6; BCBA review/sign-off labor (35 min blended, contract rate) ~$35–45; QA pass ~$5; support/follow-up amortized ~$5; rework/appeal-trigger reserve ~$5. Total COGS roughly $55–70 against a $175 average price — a gross margin in the 55–65% range before overhead. Automation %: ~40% of narrative-drafting effort automated at launch, ~70% by day 90 as the criteria library and prompt tuning mature, targeting ~85% by year one with review time as the remaining bottleneck by design (never fully automated). Cycle time: packet drafted within 24 hours of intake; delivered to client 5+ business days before deadline. Rework rate target: <8% of packets require payer-requested additional information. Escalation rate target: <5% require founder/senior-BCBA intervention. CAC payback: target under 3 months given the low-friction per-packet trial offer. Conversion assumptions: free-first-packet trial → paid subscription at ~35%; pilot → paid at ~50% given direct revenue-protection ROI. Retention: subscription model is inherently recurring — a client only churns if their client roster or authorization cadence changes.
| Channel | Why ICP reachable | First angle | Conversion assumption | Proof source | Measurement |
|---|---|---|---|---|---|
| State ABA provider associations & conferences | Owners/Clinical Directors attend for CEUs and payer-policy updates | "Free Reauthorization Readiness Scan" table/talk | 3–5% booth-to-lead | Existing vendor exhibitor presence at these events | Leads captured/scanned |
| LinkedIn founder-led content (BCBA/practice-owner audience) | BCBA owners actively discuss billing/authorization pain publicly | Teardown of a real (anonymized) denial letter and what should have been in the packet | 1–2% post-to-lead | Existing high engagement on RethinkBH/Operant Billing content on this topic S11S12 | Profile visits, DM replies, waitlist signups |
| Practice-management-vendor partner referral (CentralReach, Rethink ecosystem) | Vendors want a documentation-quality answer for customers without building it themselves | Co-marketing "reduce your denial rate" webinar | 10%+ of attendees to pilot | Vendor marketplace/partner-program precedent common in healthcare SaaS | Webinar-to-pilot signups |
| Direct outbound to independent agency owners | Owner is a named, findable, single decision-maker (NPI/state license directories) | Personalized "diagnosis" of their public reauthorization exposure (client count, payer mix) | 2–4% reply rate | Standard B2B healthcare outbound benchmarks | Replies, discovery calls booked |
Lead with a diagnosis, not a demo: "Based on your client count and payer mix, you likely manage roughly N reauthorization cycles a year and are losing an estimated $X in BCBA billable time to paperwork — want to see what a packet looks like when we produce it?" First call is a 15-minute walkthrough of one real (de-identified) sample packet, followed by a free-first-packet offer with no commitment.
The founder (or a partnering BCBA co-founder) publishes weekly short-form breakdowns of real (anonymized) payer denial patterns and what a compliant packet needs to include, building credibility as the specialist voice on this one narrow problem rather than a generalist billing consultant.
Lead magnet: the free Reauthorization Readiness Scan — the practice shares its last 3 authorization packets (or describes its current process), and receives a short gap report against payer-published criteria plus an estimate of BCBA hours currently spent per packet. Why it builds trust: immediately useful, specific, and non-generic. Pain signal captured: current denial/reduction rate, current turnaround time, current BCBA hours per packet. Follow-up mechanism: automated email plus a personal outreach offering the free-first-packet trial. Sales-ready qualification: practice has 30+ active clients, expresses a specific recent denial or reduction, and is the decision-maker.
Warm GTM starts with the founder/BCBA co-founder's existing professional network (BCBA supervision cohorts, state ABA association contacts, prior colleagues at ABA agencies), converting waitlist signups from content into consultative 15-minute diagnostic calls rather than generic demo pitches.
Outbound targets independent ABA agency owners identifiable via state behavior-analyst licensure boards and NPI registry data, personalized around each prospect's likely payer mix and estimated reauthorization volume, leading with the free Readiness Scan rather than a generic "book a demo" ask.
Publish structured, specific answers to the exact questions BCBA owners already type into search and AI assistants — "how often does ABA authorization need to be renewed," "why did my ABA authorization get reduced," "what does Optum require for ABA concurrent review" — so that both traditional search and AI answer engines surface ReauthClear as the specialist source on this narrow question, distinct from generalist ABA billing content.
Pilot cohort: capped at 5 independent ABA agencies in the first 60 days. Incentive: first packet free, first month at 50% off. Feedback mechanism: a structured post-packet survey (accuracy, turnaround, would-you-pay-full-price) plus direct BCBA-reviewer notes on every edit made to the AI draft, which feed the prompt/criteria library. Hardening rule: after 5 pilots, harden intake and evidence requirements and the QA checklist; after 10, harden SOPs, the exception queue, and reviewer checklists/delivery templates; after 20, pause new pilots until COGS, rework rate, escalation rate, and cycle time are actually measured — never scale by adding review headcount to paper over unresolved workflow gaps.
Every BCBA edit to an AI-drafted packet is logged and categorized (missing data point, wrong emphasis, payer-criteria mismatch, tone). Recurring edit categories are converted into permanent updates to the payer-criteria library and prompt instructions, so first-draft quality measurably improves with each cohort rather than staying static.
The most material risks are (1) BCBA reviewer capacity failing to scale with packet volume, mitigated by a growing contract-reviewer panel and a hard cap on new pilots until throughput is proven; (2) a payer materially changing its criteria format faster than the library is updated, mitigated by a dedicated weekly criteria-monitoring routine; and (3) data-security/BAA breach risk given the sensitivity of pediatric health data, mitigated by encrypted transfer, minimum-necessary data collection, and a documented incident-response plan. Full register below.
Likelihood: Medium · Impact: High · Mitigation: cap pilot growth to reviewer throughput; build a bench of contract BCBA reviewers before scaling client count.
Likelihood: Medium · Impact: Medium · Mitigation: weekly payer-bulletin monitoring routine; version-controlled criteria library with change alerts to affected clients.
Likelihood: Low · Impact: Severe · Mitigation: encrypted transfer only, minimum-necessary data collection, signed BAAs, documented incident-response plan, cyber liability insurance.
Likelihood: Medium · Impact: Medium · Mitigation: out-execute on speed/quality SLA and build the cross-agency criteria/outcome data moat faster than incumbents can retool.
Likelihood: Medium · Impact: Medium · Mitigation: pricing is per-packet/per-client, not per-hour-approved, so revenue is more resilient to hour-level cuts than to authorization-cycle-frequency cuts; diversify beachhead across multiple states.
Likelihood: Low-Medium · Impact: High · Mitigation: mandatory minimum review-time tracking, spot-audit a sample of signed packets, tie reviewer pay in part to audit quality, not just volume.
Likelihood: Medium · Impact: High · Mitigation: deterministic fact-checking layer cross-references every clinical claim against source session data before it reaches the reviewer; reviewer is the final control.
Likelihood: Medium · Impact: Medium · Mitigation: automated completeness check at intake with a required-evidence checklist; exception queue holds incomplete files before drafting begins.
Likelihood: Medium · Impact: Medium · Mitigation: tight, capped pilot cohort with a clear free-to-paid transition offer and a fast time-to-first-value (first packet within days).
Likelihood: Low · Impact: Severe · Mitigation: explicit disclaimers, mandatory BCBA sign-off enforced technically (no submission path bypasses it), marketing language reviewed by counsel to avoid implying independent clinical authority.
Likelihood: Low · Impact: High · Mitigation: flat per-packet/per-client pricing only, never contingent on dollar value or hours approved; pricing policy documented and disclosed to clients.
Likelihood: Medium · Impact: Medium · Mitigation: hardening checkpoints at 5/10/20 pilots enforce SOP creation before further growth.
Likelihood: Low · Impact: Low · Mitigation: use conservative ranges in planning, revisit sizing quarterly against actual pilot conversion data rather than published estimates alone.
Evidence threshold cleared: identified buyer (independent ABA agency owner/Clinical Director), painful specific recurring problem (90–180-day reauthorization cycles at 200–400/year per 100-client practice) S5, proof of existing spend (a $400–600M outsourced ABA RCM market with named competitors pricing at 5–8% of collections) S5S17, active demand evidence (payer-published treatment-plan-requirements documents, sustained trade-press content on authorization/denial pain) S10S11S12, a credible reason to win (narrow specialist wedge vs. bundled generalist RCM), a narrow MVP wedge (reauthorization-only production), a practical path to first sale (single pilot agency, no large platform build required), no unresolved fatal blocker (clean BCBA-sign-off licensing boundary, non-contingent pricing), a credible 50%+ gross-margin path, and a believable distribution path (BCBA/practice-owner communities, state ABA associations, PM-vendor partnerships). No fatal disqualifier is triggered: buyer and problem are specific, spend and demand are evidenced, licensing risk is mitigated by design, no substantial physical labor is required, and no large custom platform is needed before first revenue. Verified / Inferred
GO. Build ReauthClear as a done-for-you ABA reauthorization and medical-necessity documentation desk, launching with reauthorization-only production for a capped cohort of five independent ABA agencies, priced per packet and per active client per month — never hourly, never contingent on approved-hours value. The narrow, single-workflow wedge is the differentiator against bundled generalist RCM vendors; the mandatory BCBA review-and-sign chokepoint is the trust interface and the licensing-safety boundary; the maintained, cross-agency payer-criteria library is the long-run moat as raw AI drafting commoditizes. Start with the free Reauthorization Readiness Scan to pull warm leads from the BCBA/practice-owner community, and let the combination of rising ABA demand and simultaneously tightening payer scrutiny pull demand forward.