CodeAudible

1. One-liner

Tells audiology front desks which hearing-device code each payer and TPA actually accepts this week, before the claim goes out.

2. Trend signal — why now?

On 1 January 2026, the AMA deleted CPT codes 92590–92595 — the code set audiologists had used for hearing aid services for decades — and replaced them with 12 new codes, 92628–92642, covering candidacy evaluation, selection, fitting, follow-up, verification, and assistive device services. Eight of the twelve are time-based, governed by the “half plus one” rule: a 30-minute code needs at least 16 minutes documented, a 15-minute code at least 8.

That alone would be a one-time retraining problem. Three things make it an ongoing revenue leak instead:

One — there is no national fee schedule to anchor to. The RUC recommended carrier pricing, so the new codes carry no assigned RVUs. Payment is determined entirely by individual payer policy and provider contracts. And because hearing aids and related exams are statutorily excluded from Medicare, CMS confirmed all 12 codes stay non-payable for CY 2026. There is no CMS rate card to copy. Every practice negotiates and codes blind.

Two — payers did not migrate in lockstep. ASHA’s own guidance is explicit: “some payers may not immediately replace the deleted codes or may elect to continue using existing HCPCS codes for hearing aid services,” and practices must “check directly with payers, including state Medicaid programs and third-party payers, before using the new codes.” The V5 HCPCS codes did not change at all, so a third parallel code set stayed live. Some carriers have been slow to update their systems, which drives denials.

Three — the professional bodies have no answer. The American Academy of Audiology’s own FAQ concedes “it often takes payers some time to adopt new CPT codes and activate them in their systems” — and then provides no timeline expectation, no fallback coding strategy, no procedure for verifying whether a specific payer has activated the codes, and no denial-handling guidance. The Academy of Doctors of Audiology withdrew from the joint code workgroup in 2023 and published “New CPT Hearing Aid Service Codes: A Cause for Concern,” citing “considerable concerns regarding the establishment of these codes, the elimination of codes 92590-92595, and the potential negative impact of these actions on private practices.”

Result: for the same fitting appointment, the correct code depends on who is paying — new CPT, legacy CPT, or V5 HCPCS — and nobody publishes the answer.

Provenance:
  - Signal 1 (Demand): ASHA guidance confirms payers split across three live code sets and instructs practices to verify each payer individually before billing; denials driven by deleted-code use and missing total time — https://www.asha.org/practice/reimbursement/coding/new_codes_aud/ — observed 2026-08-27
  - Signal 2 (Feasibility): AAA FAQ confirms no RVUs, carrier pricing, and no published guidance on payer adoption status, fallback coding, or denial procedure — the gap is unfilled by the professional bodies themselves — https://www.audiology.org/practice-resources/coding/hearing-device-services-codes-resource-center/faqs/ — observed 2026-08-27
  - Signal 3 (Economic): 57% of private-practice audiologists work with third-party administrators, each with its own contracted fee schedule and code rules; ASHA 2023 Audiology Survey, n=439 — https://www.asha.org/siteassets/surveys/2023-audiology-survey-private-practice.pdf — observed 2026-08-27
  - Signal 4 (Economic): CMS finalized all 12 new codes as non-payable for CY 2026 under the Medicare statutory hearing aid exclusion, removing any national rate anchor — https://www.audiology.org/cms-finalizes-cy-2026-physician-fee-schedule-key-takeaways-for-audiology/ — observed 2026-08-27
  - Signal 5 (Demand): Industry billing commentary reports insurance verification "sometimes taking an entire morning on hold" and administrative tasks consuming over half of clinician time — https://www.allzonems.com/audiology-practice-challenges/ — observed 2026-08-27
  Category: Regulatory arbitrage

3. The opportunity

The incumbents solve an adjacent problem and miss this one cleanly.

AAPC Codify and OptumCoding sell code crosswalks — 18 crosswalk types, ICD-10-PCS to CPT, CPT to HCPCS, updated quarterly. They map code to code. They do not map code to payer. Knowing that 92590 was replaced by 92628 is the easy half; knowing that Blue Cross of your state still adjudicates 92590, that a given TPA wants V5011, and that a third payer activated 92628 in March is the half that gets you paid.

CrossCoder indexes active and retired CMS policies for claim validation. For hearing device services that is structurally useless — Medicare doesn’t pay for them at all. The one thing it’s built on is the one payer that will never be relevant here.

Sycle, Blueprint OMS, and CounselEAR own the practice-management workflow and will absolutely add the new code numbers to their pick-lists. That is table stakes and I assume they’ve done it. What a PM system will not do is maintain a living, per-payer, per-state adoption record — that’s a research operation, not a software feature, and it doesn’t fit their business model.

Billing services (5–8% of collections, or $3–10 per claim) will handle it for you, but they price for practices with real claim volume and they take a permanent cut of revenue. A two-audiologist practice doing mostly private-pay dispensing with a handful of TPA contracts is exactly the customer that can’t justify that.

The gap: a maintained matrix of which hearing-device code each payer and TPA accepts, by state, updated as adoption moves — checked at the point of claim. It’s a data-collection business wearing a SaaS interface.

4. Target market

• Primary customer: Owner-audiologist of an independent US hearing care practice, 1–4 audiologists, single or two locations, that bills at least three payers or TPAs. The person who buys is the owner; the person who uses it daily is the front-desk/billing coordinator.
• Why they buy: Every denied fitting claim is a rework cycle — re-code, re-file, chase, sometimes write off. Verification calls already burn staff hours (“an entire morning on hold”). The 2026 transition turned a settled routine into a per-payer guessing game, and the professional bodies explicitly told them to figure it out themselves.
• Rough TAM reasoning: I’ll be honest that this number is soft. IBISWorld counted 11,783 US hearing aid stores with employment in 2025, but that bucket mixes independents with Costco, chains, and franchises. The best independent-practice estimate I could source is 2,760 (23% of ~12,000 outlets) — and that data is from 2011, with a documented consolidation trend since. ASHA’s 2023 private-practice survey drew on 439 respondents, 184 of them owners, from a certified-audiologist population of ~11,000. Realistically the serviceable base is 3,000–6,000 independent practices. That’s small — which is precisely why it’s unserved, and it’s enough for a $1–3M ARR business, not more.
• Why now for them: The code set changed nine months ago. Practices are now far enough in to have accumulated denials, and adoption is still visibly in motion — which is the window where a live matrix has value. In two years the payers converge and this decays into a static reference.

5. Product sketch (MVP)

• Payer matrix lookup — pick payer/TPA + state + service, get the code set that payer is currently adjudicating (new CPT / legacy CPT / V5 HCPCS), with a last-verified date and the evidence behind it.
• Pre-submission claim check — paste or upload a claim line; flags deleted-code use, missing total time on time-based codes, and modifier problems before it goes out.
• Time-threshold calculator — enter documented face-to-face minutes, get the correct code and unit count under the “half plus one” rule so 92628/92629 add-on stacking is right.
• Denial decoder — upload the remit/EOB, get a plain-language reason and the specific corrected resubmission, and the outcome feeds back into the matrix.
• TPA contract shelf — store each TPA’s contracted fee schedule and code requirements side by side, so the front desk stops digging through PDFs and emails.
• Change alerts — email the practice when a payer they actually bill flips its code policy.
• Verification call script — generates the exact question list for a payer rep, and a one-click way to log the answer back into the matrix.

6. AI angle — what’s load-bearing

Remove the AI and this is a wiki that goes stale in six weeks. AI is doing two jobs that are otherwise unaffordable at this market size.

Ingestion at the edges. Payer adoption evidence arrives as garbage: PDF provider bulletins, remittance advice, EOB scans, portal screenshots, TPA policy attachments. Document extraction turns that mess into structured matrix rows — payer, state, code accepted, code rejected, date, source. This is what makes a 3,000-practice market economically viable to serve; a human research team maintaining this by hand costs more than the market pays.

Denial-to-evidence loop. Each uploaded denial is a live observation of one payer’s real behavior on one date. The model reads the remit, classifies the cause, and updates the matrix. Every customer’s rework makes the product more accurate for everyone else — the compounding asset here.

What is deliberately not AI: the matrix answer itself. A hallucinated billing code creates a denial, which is the exact harm we’re selling against. Every cell shows its source document and verification date, and the model’s job is to extract and route evidence, not to guess codes.

7. Localization angle

N/A — this is a US-only play. The entire product is a function of the AMA CPT code set, CMS carrier-pricing rules, the Medicare hearing aid statutory exclusion, state Medicaid programs, and the US hearing-TPA cartel. None of that structure exports. Any other geography would be a ground-up rebuild against unrelated rules, not a translation.

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

• Pricing: $149/mo single-location practice; $299/mo multi-location or 3+ providers. Annual prepay at 2 months free.
• ACV: ~$2,100 blended.
• Rough math to $1M ARR: ~475 practices at blended ACV. Against a serviceable base of 3,000–6,000, that’s 8–16% penetration — demanding but not fantasy for a category with no direct competitor.
• Rough math to $5M ARR: Not credible on this customer base alone. $5M needs ~2,400 practices — 40%+ of the entire independent market. The honest ceiling here is $1.5–3M ARR. Getting past that means expanding the same payer-matrix engine into adjacent small specialties with the same structural problem (optometry vision plans, DME, orthotics) — a real path, but a second product decision, not an upsell.
• Expansion path: Per-location fees as practices add sites; a per-claim denial-recovery tier for practices that want the resubmission drafted; and eventually selling anonymized adoption benchmarks back to state academies.

I’d rather write down a truthful $2M business than dress this up as a $5M one. For a solo operator, $2M ARR at software margins on a market nobody else wants is a good outcome.

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

• State academy circuit. Audiology runs on state academies (Florida Academy of Audiology and peers already publish 2026 coding-change explainers — they are actively looking for content on exactly this). Offer a free 45-minute “2026 code transition — what your payers are actually paying” webinar to 10 state academies. Bring real matrix data for their state’s payers. This is the highest-intent room that exists for this product.
• Denial teardown as the hook. Cold-email independent practices offering a free audit: send 10 denied hearing-device claims, get back a coded analysis of what each payer wanted. It’s a working demo, it costs one AI run, and it collects the exact evidence that builds the matrix. Target list is buildable from state hearing aid dispensing board licensee registries, which are public.
• ADA and AAA member channels. The ADA has publicly declared “considerable concerns” and published member resources on these codes. A practical tool that answers the question their own FAQ leaves open is a natural fit for member-benefit placement or conference sponsorship.
• Coding consultant partnerships. A small number of named consultants (Kim Cavitt is the recurring name in ADA’s code resources) are the profession’s trusted coding voices. One affiliate arrangement reaches more practice owners than any ad spend.
• Seed the matrix before selling it. Pre-launch, run verification calls on the top 20 payers/TPAs across the 5 largest states. That’s a few weeks of phone work and it converts the product from empty database to day-one useful.

10. Build complexity — justification

Medium. The software is unremarkable — a structured matrix, document extraction, a rules engine for time thresholds, alerting. Off-the-shelf models handle the PDF/EOB ingestion. No integration with practice management systems is needed for v1; paste-and-check is enough to prove value, and avoiding Sycle/Blueprint integration removes the hardest dependency.

The genuine work is data acquisition: making the first few hundred verified payer×code cells real, by phone and by document, before any customer arrives. Call it 10–14 weeks to v1 for one builder plus a part-time domain person, with the domain person’s calling time being the actual critical path. That domain requirement is real and I’ve scored build feasibility down for it.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Billing guidance tool. Publishes verified payer behavior; doesn’t submit claims or give legal advice.
Ethical — no harm / dark patterns	✅	Reduces denials and rework. Every cell shows source and date; no guessed codes.
Market exists (evidence above)	✅	Code set deleted, three parallel sets live, professional bodies concede no guidance exists.
1–5 person team can build this	✅	One builder + one domain-expert audiologist/coder.
Launchable with <$50K / ₹40L	✅	Inference, hosting, and the domain expert’s contract hours. Well under $50K.

All five pass.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	16/20	Real, recurring, costs money on every misrouted claim, and the bodies they trust told them to solve it alone. Not quite hair-on-fire: many practices dispense largely private-pay, so it bites hard for TPA/insurance-heavy practices and only moderately for others.
Demand evidence	15	11/15	Strong documentary evidence — ASHA, AAA and ADA all confirm the gap in writing, plus 57% TPA participation. Docked because I could not source verbatim practitioner quotes; audiology forum discussion isn’t publicly indexed. The pain is inferred from institutional sources, not heard directly.
Build feasibility	15	12/15	Standard stack, off-the-shelf extraction, no PM-system integration needed for v1. Held back by the manual data-seeding required before launch.
Distribution clarity	15	11/15	State academies and licensee registries are named, finite, reachable channels with high intent. Not a 2-week sprint — webinars and academy calendars run on their schedule, not yours.
Revenue mechanics	15	11/15	Pricing sits comfortably below a 5–8%-of-collections billing service. Docked because the customer base is genuinely small and $5M isn’t reachable within this niche.
Time to first revenue	10	7/10	Denial-teardown offer converts warm; realistic first paid customer in 6–8 weeks post-launch, but the matrix must be seeded first.
Defensibility	10	4/10	The accumulating verified-adoption dataset is a genuine 12-month moat and gets better with every customer denial. But it’s copyable by anyone willing to do the phone work, and — the real risk — the moat decays as payers converge on the new codes.
Total	100	72/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · domain-expertise-required

This does not work without a co-founder or first hire who has actually billed audiology claims. The matrix is only credible if someone who knows the difference between a V5011 and a 92634 is verifying it. A generalist builder will produce a plausible, wrong database — the worst possible outcome for this product.

Key assumptions to validate (3–5)

1. Assumption: Payer adoption is genuinely fragmented — meaningfully more than one code set in live use across major payers/TPAs today. How to test: Call 20 payers/TPAs across 5 states and record which code set each adjudicates. If ≥80% have converged on 92628–92642, the core premise is gone.
2. Assumption: Independent practices are experiencing real denials from this, not just anxiety about it. How to test: Run the free denial-teardown offer with 25 practices. Count how many actually have hearing-device denials traceable to code-set mismatch.
3. Assumption: Practices will pay $149/mo for lookup rather than absorb it as staff time. How to test: Pre-sell annual at a founder discount to 15 practices before building. Fewer than 5 closes means the pain is being tolerated, not bought.
4. Assumption: The independent practice base is ≥3,000 — not the 2,760 of 2011 further eroded by 15 more years of consolidation. How to test: Count active licensees across 5 state dispensing boards and extrapolate; cross-check against ASHA membership data.

Risk flags

1. Market timing — narrow and closing. This is the strongest risk. The value exists because payers are mid-migration. As they converge on 92628–92642, the matrix flattens into a static reference anyone can publish. The window is realistically 18–30 months, and the business must convert into stickier ground (denial recovery, TPA fee benchmarking, adjacent specialties) before it shuts.
2. Small TAM with real uncertainty. The serviceable base may be closer to 3,000 than 6,000, and the only hard independent-practice count I could find is 15 years old. If it’s at the bottom of that range, $1M ARR needs ~16% penetration — hard.
3. Incumbent absorption. Sycle/Blueprint/CounselEAR could bundle a payer matrix into the PM system they already sell. I judge it unlikely — it’s ongoing research labor, not a feature — but they own the workflow and the customer relationship.
4. Data liability. A wrong cell causes a denial, which is the harm we sell against. Requires visible sourcing, verification dates, and conservative “unverified” states rather than confident guesses.

14. Structured verdict

Score:                  72/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical solo founder paired with a practicing audiology biller/coder as co-founder or first hire
Time to revenue:        6–8 weeks post-launch; ~4 months from start including matrix seeding
Capital to launch:      $12–18K (₹10–15L) — mostly domain-expert contract hours and inference
Top 3 assumptions to validate first:
  1. Payer fragmentation is real — call 20 payers/TPAs across 5 states, record live code set; kill if ≥80% converged
  2. Denials are actually occurring — free teardown with 25 practices, count code-mismatch denials
  3. Willingness to pay $149/mo — pre-sell 15 annual seats before writing the product
Kill criteria:
  - Abandon if ≥80% of surveyed payers/TPAs have already converged on 92628–92642 (the fragmentation is the product)
  - Abandon if fewer than 5 of 15 pre-sell targets convert at $149/mo
  - Abandon if Sycle or Blueprint OMS ships a maintained per-payer adoption matrix before v1
  - Abandon if the verified independent-practice base proves to be under 2,500

15. Next step — 1-week validation sprint

• Day 1–2: Build the target list from public state hearing aid dispensing board registries in FL, TX, OH, CA, NY. Separate independents from chains and franchises. This simultaneously tests the TAM assumption and produces the outreach list.
• Day 3–4: Make 20 verification calls — 12 commercial payers, 8 TPAs (TruHearing, UnitedHealthcare Hearing, NationsHearing, HearUSA, Hearing Care Solutions, Amplifon among them) across those 5 states. One question: which code set are you adjudicating for hearing aid fitting and follow-up today? Record every answer with a date. This is both the validation and the first 20 rows of the product.
• Day 5: Offer the resulting mini-matrix free to 25 practice owners by email in exchange for a 15-minute call about their denials, and pitch a founder-rate annual prepay at the end of each call.

Falsifiable outcome: Go only if (a) at least 5 of 20 payers/TPAs are on a different code set than the others — proving live fragmentation — and (b) at least 3 of the 25 practices commit money at founder rate. If payers have converged, the product has no reason to exist and I stop; if they’re fragmented but nobody pays, the pain is being absorbed by staff time and I stop. Either way I know inside a week, for the price of a phone bill.