What changed
FACT (Guardian, id 6169): a rising wave of ADA web-accessibility demand letters and serial-plaintiff lawsuits is hitting US cafe, shop and small e-commerce owners, who report frustration and no cheap remediation path. HYPOTHESIS: LLMs can now reliably convert raw automated-scan violations (id 6436) into ranked, human-readable fixes with code snippets — the piece that was previously manual consultant work.
Why now
FACT: litigation volume is rising and owners are actively searching for cheap remediation. The panic-buy window opens the day a demand letter arrives — an acute, card-in-hand moment. INFERENCE: overlay incumbents (accessiBe) have been publicly criticized and even sued for NOT actually delivering compliance, opening room for an honest 'fix the code' alternative.
Converging signals
Complaint/pain signal (ADA lawsuits frustrating owners) × AI capability (LLM turns scan output into ranked remediation). A pain-x-capability convergence, NOT a government forced-filer mandate — there is no portal to submit to.
Customer pain
FACT: owner receives a demand letter threatening statutory damages + attorney fees, has no in-house dev, doesn't know what's broken or what 'good faith' looks like, and the alternatives are a $49/mo overlay widget of dubious legal value or a several-thousand-dollar consultant/attorney.
Who pays
A local cafe/retail/small e-commerce owner (or their web person) who just received or fears a demand letter, paying by card this week to reduce legal exposure. Discretionary but acute buyer.
Solved today
Free scanners (WAVE, axe DevTools, Lighthouse) that dump raw violations they can't interpret; overlay widgets (accessiBe, UserWay, AudioEye) at $29-$99/mo; or expensive accessibility consultants / ADA attorneys.
Why current solutions are bad
Free scanners produce unranked jargon with no remediation or legal artifact. Overlays are widely reported to NOT stop lawsuits (plaintiffs specifically target overlay sites) and give false security. Consultants are slow and cost 10-50x. None hand the owner a defensible, WCAG-mapped good-faith-remediation record.
Proposed product
URL in → run axe-core + Lighthouse a11y scan → LLM ranks violations by legal risk + effort, writes plain-English fixes with copy-paste code snippets → generates a WCAG 2.1 AA-mapped 'good-faith remediation plan' PDF timestamped for the owner's attorney. Explicitly NO overlay widget. Upsell: $29/mo re-scan + change monitoring.
MVP version
Single-page app: paste URL, queue a headless scan (axe-core/Playwright + Lighthouse), pipe JSON through an LLM prompt that outputs ranked fixes + snippets, render a branded PDF. Stripe checkout at $149. Buildable solo in 1-2 weeks on cheap infra.
30-day build
Ship the scanner→report→PDF pipeline and Stripe. Harden the LLM prompt against hallucinated fixes (cite the exact WCAG success criterion and the DOM node for every item). Hand-test on 20 real small-business sites; get 3-5 paid pilots via direct outreach to owners named in recent filings and to the web designers/agencies that build for them.
60-day build
Add re-scan monitoring subscription, a white-label mode for freelance web designers/agencies (recurring channel), and a lawyer-referral angle (ADA-defense attorneys need a fast, credible remediation vendor to point clients to).
90-day revenue plan
Two channels producing revenue: (1) direct panic-buy audits at $149, (2) designers/agencies reselling audits + $29/mo monitoring to their whole client book. Target first meaningful revenue inside 30-90 days given the acute buyer.
Distribution path
Direct outreach to businesses in public ADA-lawsuit dockets/demand-letter patterns; partnerships with web-design freelancers/agencies (they own the buyer and fear liability too); content/SEO on 'ADA demand letter what to do'; affiliate/referral with ADA-defense attorneys.
Pricing hypothesis
$149 one-time audit + remediation PDF; $29/mo re-scan & monitoring; agency white-label tier (e.g. $99/mo for bulk client scans). Undercuts consultants, priced above a throwaway widget because it delivers an actual legal artifact.
Technical difficulty
Low-moderate. axe-core/Lighthouse are mature open-source; the LLM report layer and PDF generation are straightforward. Main risk is prompt reliability — must not invent fixes; ground every item in the actual scan node + a named WCAG criterion.
Legal / regulatory risk
Moderate and must be handled carefully: automated scans catch only ~30-40% of WCAG issues, so the product must NOT claim 'compliant' or 'lawsuit-proof' — position it as a good-faith remediation aid, not legal advice, with clear disclaimers. Overstating protection is the real liability. Founder does not need a license to sell a scan+report (does NOT constitute practicing law).
Platform dependency
None material — no app-store or government-portal gatekeeper. Depends on open-source scanners and an LLM API (swappable). No platform can deplatform it.
Founder fit
Moderate. Fits the founder's micro-SaaS / compliance-monitor / complaint-mining wheelhouse and AI-workflow strength, and value is sold by demonstration (paste a URL, see the report). But it is NOT his proven government-portal forced-filer shape (no mandate, no per-filing portal), so founder-fit is good, not maximal.
Breakout potential
Moderate. Could expand into a recurring monitoring product and an agency white-label channel across the huge long tail of small-business sites. Ceiling is capped by the overlay incumbents' distribution and by buyers' difficulty distinguishing an honest fix-list from a marketed widget.
Final recommendation
BUILD-SMALL / VALIDATE. A legitimate acute-pain quick win with fast, cheap MVP and card-today buyers, but the moat is thin and incumbents own distribution. Green-light a 2-week MVP and prove the wedge by landing 3-5 paid audits through the agency/attorney channel before scaling spend. Lead with the legal-defense PDF + honest 'fix the code, not a widget' positioning, never a compliance guarantee.
Next action
Build the URL→axe/Lighthouse→LLM-ranked-report→WCAG-mapped PDF pipeline with Stripe at $149, then cold-outreach 20 web-design freelancers/agencies and 10 owners named in recent ADA web filings to sell the first five audits.