Plan vs. Bill: Upload Your Plan, Snap the Bill, See the Truth
The product is the app. The moment is the chair. The moat is distribution and trust — not the AI. This is the corrected direction after the first deep dive pointed at the wrong target.
Verdict
The app is the business. Pain point + easy solution + savings beats any model. The first deep dive aimed at the wrong target.
What the first deep dive got wrong: it concluded "build the system, not the app" — content and a plan database with the tool demoted to a lead magnet. That underweighted the thing that actually decides winners in consumer products: being present at the moment of pain, with a working answer, in seconds. Nobody is problem-aware in the dental chair. Whoever intercepts sticker shock wins, and that is a distribution and workflow fight, not a model fight.
What stands: the gap is real, dental is the right beachhead, the plan database still matters as data depth. But they orbit the app. The app is the product.
Verdict: build the app Dental first Distribution is the moat Workflow beats models
The Chairside Moment
This product wins in one place: the 60 seconds where a patient holds a written treatment plan with prices on it and has to decide. That moment is unique to dentistry — medicine bills you after; dentists hand you a quote before. Nobody opens ChatGPT there. Nobody has ever lined up the plan document against the treatment plan. The app shows up at that moment and hands over a decision, not a paragraph.
Step 1 — Upload once
Insurance plan PDF or card photo. We parse: annual maximum, deductible, tiers (100/80/50), frequency limits, downgrade clauses ("posterior composites paid at amalgam rate"), waiting periods, missing-tooth clauses.
Step 2 — Snap in the chair
Photo of the written treatment plan: procedure codes, tooth numbers, prices. No typing. Works at the front desk while the estimate is in your hand.
Step 3 — Get the one screen
Minimum that holds up · upgrade ladder with real dollar deltas · your true cost against the plan · three questions to ask. A decision, not a document review.
Step 4 — Share the result
"I saved $640 — here's what I asked." The output card is screenshot-able and shareable. The product's results are its best marketing.
The product law: never bash the dentist
✅ What the app says
"The lab crown is the standard of care and costs less. The same-day crown buys convenience. Here is the exact delta under your plan, and three questions to ask so you can decide."
❌ What the app never says
"You're being scammed." "Don't get the crown." "Your dentist is ripping you off." The app is informational — not medical or insurance advice — and every output carries that disclaimer.
Sample Output — What the App Actually Returns
Illustrative example: porcelain crown, tooth #14, quoted at $1,800. Plan: PPO, 50% coverage on major services, $1,500 annual maximum, downgrade clause on posterior tooth-colored fillings.
🦷 Crown options under YOUR plan
This is the entire value in one screen. The math is deterministic against the parsed plan; the explanation is plain English. Not a bash on the recommendation — a map of the choices.
Messaging & Brand
The brand voice has two registers, and keeping them separate is a product law, not a style preference.
📣 Ad and social voice — sharp
This is where the hooks live. Corporate dentistry is the target of the message: private-equity rollups, production quotas, the upsell baked into the business model. The dentist is often a fellow captive of that system, not the villain.
🧾 Product and site voice — professional
Always credible, calm, evidence-first. "Options and reality." The marketing can be loud; the analysis must be trustworthy. The trust is the product. Disclaimers on every output.
Approved hook direction
Domain and naming
mybilldoctor.com RECOMMENDED
Says exactly what it does. Works beyond dental (matches the expansion play). Passes the radio test. Watch: "doctor" implies authority — the "informational, not medical or insurance advice" disclaimer layer becomes more important with this name.
mybilldr.com REGISTER DEFENSIVELY
Good short registration to hold alongside the primary. "Doctor" spelled out is what people will say out loud and type by default.
"Plan vs. Bill" can remain the product/feature name or the content brand; the company brand is the advocate. Naming is reversible — the trust you build in it is not.
Why Dental Wins First
Written prices before you decide
Dentists hand you a treatment plan with dollar amounts before any work happens. That quote-in-hand moment is the entire product surface. Medicine bills after the fact — the hook barely exists there.
Discrete, comparable upgrades
Amalgam vs. composite. Metal vs. porcelain vs. zirconia. Same-day vs. lab crown. Dental upsells are concrete material choices — exactly where "marginal difference, save $X" analysis lives.
Plans are complicated but parseable
Annual maximum, 100/80/50 tiers, frequency limits, downgrade clauses. Opaque enough that nobody reads them, structured enough to be solved. A wedge, not a swamp.
The money is real
Consumers pay cash at the counter, cost-per-click on dental problem keywords runs $5–16, and the anxiety is immediate. Nobody negotiates with a dentist mid-appointment — they pay and regret later. The app breaks that pattern.
Competition: Who Is Actually In This Space
| Player | What they do | Compares to YOUR plan? | Threat |
|---|---|---|---|
| Goodbill | Reviews hospital bills, negotiates reductions (human, fee on savings) | No — fair pricing | LOW Different business |
| Resolve | Advocate audits and negotiates hospital bills | No — negotiation | LOW Could refer to you |
| HealthScan AI | Scans bills, compares charges to fair prices | No — market reference point | MEDIUM Closest neighbor |
| DecodeMyForm AI | Explains medical bills and EOBs in plain English | No — one document | MEDIUM Half the feature |
| Counterforce Health | Analyzes bills, fights denials | No — denial focused | MEDIUM Downstream of your moment |
| Codify / ICD10data | Code lookup for professional coders | No — reference only | HIGH on SEO Owns code keywords |
| Cassidy AI | Compares plan renewal quotes (B2B brokers/HR) | Yes, but broker side | LOW Validates B2B demand |
| ChatGPT / Gemini | Anyone can upload both documents and ask | Yes, if the user thinks to ask | HIGH See AI section |
Read of the field: the industry organized around two moments — after the bill hurts (negotiation, denial fighting) and for professionals (code tools). The moment in the middle — "I have a treatment plan in my hand, what will this actually cost me under my plan?" — is unoccupied. That is the wedge, and the app is built for exactly it.
The AI Question — Why "ChatGPT Kills It" Is the Wrong Fear
The objection: anyone can upload both documents to ChatGPT and get a comparison for free, so the app is dead on arrival. Here is why that overstates the threat — and where the real threat actually hides.
✅ The prompt gap is real
AI is pull-based. It only helps people who know what to ask. Nobody in the chair thinks "let me audit my plan's downgrade clause." Capability is not access. People do not prompt their way to an answer they cannot conceive of.
✅ Workflow beats models
The app's moat is 30 seconds: snap, decision, done. Generic AI gives a paragraph if the user does the homework — uploads the plan, reads the clause, frames the question. The product has already returned the answer.
⚠️ But the gap cuts both ways
The same unawareness that stops people from prompting AI also stops them from finding the app. The prompt gap is not protection — it is the acquisition problem. Whoever solves awareness wins, and that is solved with distribution, not technology.
⚠️ The real threats
1. Carriers ship plan-explainer AI in their own apps. 2. Fast followers ship the same feature once parsing is commoditized (~18 months). 3. AI assistants add memory: "I already have your plan" neutralizes the plan-reference argument over time.
Distribution: The Actual Moat Maker
Content, social, and newsletters are not marketing around the product — they are the weapon that makes the product work. The game: catch the person at sticker shock, before or after the appointment, and turn "is this normal?" into "let me snap it."
Social video
Real treatment plan in, real analysis out, on screen. The format is inherently dramatic: the quote, the reveal, the savings. Nothing about it needs to be explained twice.
Newsletter
Retention engine and the brand voice home. "What a crown should cost," "the upgrade clause your plan is hiding." Cheap to run, compounds trust.
SEO content
"Dental crown cost with insurance" — 2,400 searches a month, near-zero difficulty. Consumer problem keywords ($5–16 CPC) are uncontested because professional coders never search them and consumer sites never built for them.
The shareable output loop
Every result is a marketing asset: "saved $640, here's what I asked." Screenshots travel. The product's output feeds the top of the funnel — the loop closes itself.
Content's job is not to teach. Informational content is exactly what free AI answers eat. The content exists to make someone say "wait — there's a tool that checks MY plan?" and hand over their documents. The call to action is always the snap, never the read.
Revenue
| Path | What it is | When | Why it works |
|---|---|---|---|
| Freemium tool | Free first check; subscription for saved plan, household vault, unlimited checks, renewal alerts ($7–10/month) | Month 6–12 | The free product is the marketing. Do not lead with the paywall |
| Insurer / employer sponsorship | Carriers or self-funded employers fund the tool as a member benefit | Month 6+ | Members who are not overbilled or over-upgraded file smaller claims — direct savings for the payer. The "insurer buys the podcast for exposure" logic |
| Content traffic + affiliate | Dental savings plan referrals, advocacy referrals, broker leads on problem keywords | Month 3–9 | CPC data ($5–16) proves advertiser density. Illustrative: 20,000 visits at 2% referral and $15 per action ≈ $6,000/month |
| Plan data licensing (B2B) | Structured, verified plan tables to brokers, HR tech, dental offices | Year 2+ | Cassidy AI already sells plan comparison on the broker side — the spend exists |
Expansion: Brand, Not Roadmap
The vision is bigger than dental and should be — from day one the brand is "help yourself against a system built to overbill and upgrade you." Bills that make people say "wtf, this is too much" — electric, gas, insurance, any opaque quote — are the eventual territory.
What expands now
The voice, the newsletter, the audience, the trust. "My Bill Doctor" as a name is bill-agnostic from the start, so the brand can grow without a rebrand.
What does not expand yet
The product. Every "snap a bill" vertical has different data, different clauses, different liability. Dental proves the model first. The second product is dental post-bill and EOB review; everything after that waits for dental to win.
Build Plan — What Codex Builds First
1. Landing page
The promise, the sample output card, an email capture. Ships before the engine works. CF Pages, existing template.
2. Plan parser
Plan PDF to structured benefits (the hard part): annual max, deductible, tiers, frequency limits, downgrade clauses, waiting periods. This is where the real engineering lives.
3. Treatment plan capture
Photo to procedure codes, tooth numbers, prices. Optical recognition with human-checkable results.
4. The analysis engine
Deterministic math for cost rollup against the parsed plan. Language model only for the plain-English explanation and the questions-to-ask. Never let the model do arithmetic on someone's bill.
5. The shareable output card
Generated image: what was quoted, what it should cost, what to ask. Built to be screenshotted. This closes the growth loop.
Sources & Notes
Keyword volumes and CPC figures carried over from the first deep dive's research. Sample output prices are illustrative. This report is informational analysis of a product concept — not medical, dental, or insurance advice.