Auto Warranty Claim-Truth Database

A crowdsourced, non-affiliate database where actual US vehicle service contract holders log real claim outcomes (approved/denied, payout, denial reason, days-to-resolution) by provider and repair type, replacing the affiliate-funded "best extended warranty" review sites that profit from referring buyers to the same low-rated companies, monetized via a one-time pre-purchase report, data licensing to consumer-protection researchers and plaintiff firms, and a no-referral "verified good-actor" badge for above-average-approval providers.

研究阶段进度

① 需求扫描

② 市场调研

③ 可行性分析

分诊打分总分: 27/35

市场调研评估

7.5/10

评估阐述

Scoring basis: demand side (0-5, urgency/evidence quality/size and growth of affected population) plus competition side (0-5, gap confirmation/execution and legal risk/differentiation durability), summed and expressed on a 0-10 scale.

Demand side (4.3/5): Independently re-verified via direct fetch: the FTC mailed $9.6M in refunds to 168,179 consumers on Dec 11, 2025 closing out CarShield's deceptive-claims-handling case, five weeks before this research pass. A second major provider, Endurance Warranty Services, is separately defending an unresolved March 2025 federal class action alleging the identical claim-denial pattern, confirming this is category-wide, not one bad actor. Two independent complaint aggregators (pissedconsumer.com, automoblog.com/BBB) both surface first-person denial stories; Consumer Checkbook, the one credible non-affiliate voice, was directly re-confirmed to have given up on ranking providers rather than build the comparison tool this opportunity proposes. Average US vehicle age hit a record 12.8 years in 2025 (289M vehicles on road), continually growing the pool of VSC shoppers. TAM is population-derived at roughly 30-40 million US consumers/year shopping for or holding a VSC (derived from vehicle sales, price-range, and industry-revenue data, not a single reported statistic, flagged medium confidence given wide variance across market-research firms); SAM roughly 5-8 million/year reachable via search, forums, and the active FTC/litigation news cycle; Year 1 SOM modeled bottom-up at roughly $30K-$110K in combined report-sales, data-licensing, and badge-fee revenue, a cold-start scenario, not a vendor forecast.

Competition side (3.2/5): The specific product, a structured, submission-based claim-outcome ledger, has a confirmed gap: no provider, broker, affiliate site, or nonprofit runs one today, verified directly against Consumer Checkbook. A concrete B2B buyer for licensed data already exists (FeganScott's active Endurance litigation). Held below a higher score by: (1) cold-start is unproven, whether past VSC buyers will submit real denial letters and payout figures at volume is not yet evidenced for this category; (2) legal/moderation overhead from naming named companies by claim statistics is real, illustrated directly by the BBB's 2010 pay-to-play scandal and ConsumerAffairs' RICO litigation over paid-accreditation influence on review treatment, both close analogues to this opportunity's own 'verified good-actor badge' revenue line and a risk that must be actively managed, not just disclaimed; (3) the affiliate-content field, while low-trust, is deeply SEO-entrenched for the exact search terms a shopper would use. Net: strongest demand-side evidence this research process can produce at this stage, tempered by real execution and legal risk rather than any doubt about the underlying pain point.

可行性评估

淘汰

可行性评分4.8/10

评估阐述

Scale 0-10 (one decimal), 0 = no realistic path to a working business, 10 = clearly buildable, well-financed and low-risk. Weights financial viability, risk severity, compliance posture and execution difficulty together. Calibrated against the same crowdsourced outcome-database method used elsewhere in this library.

What holds the score up. Demand evidence is the strongest in this category: a $9.6M FTC refund distribution to 168,179 consumers mailed Dec 11 2025, a second major provider (Endurance) still defending an unresolved March 2025 federal fraud class action, a record 12.8-year average US vehicle age across 289M vehicles, and a competitive gap confirmed directly against the one non-affiliate voice in the category. Technical risk is near zero. Absolute capital is modest at $62,600. Federal and state complaint-index precedent (CFPB, NAIC) legitimises the concept, and Section 230 plus Kimzey v. Yelp (9th Cir. 2016, aggregate ratings from a neutral tool remain third-party content) give the core data product defensible legal ground.

Why it lands below the pass bar. No revenue line covers its own acquisition cost. A $15 report nets $13.84 and is bought once, giving LTV $14.53; paid CAC runs $69 (best case, $3.46 CPC at 5% conversion) to $195 (automotive benchmark $3.90 CPC at 2%), so LTV/CAC is 0.07 to 0.21 against a 3.0 bar. Paid acquisition is structurally impossible because affiliate programmes in this category pay $100-200 per referred warranty lead, setting an auction floor roughly 10x above what this product can bid. Organic-only CAC ($2.67 to $30.00 across the 30K-75K first-year traffic range) clears 3.0 only in the best corner. Steady-state fixed cost is $25,600/yr; report sales alone break even at 1,762 reports (88,100 visits at 2% conversion), reachable only at the extreme top of the modelled range. Payback on $62,600 is 3.2 years in the base case and effectively never in the conservative case.

Cold start (stress-tested, HIGH). The venting population is large (CarShield 51,000+ Trustpilot reviews, 2,900+ BBB customer reviews; Endurance ~3,666 BBB complaints over three years, secondary-sourced) but every one of those is friction-free text on a platform that already has traffic. This product requires a document upload plus structured payout and resolution fields, under its own rule that undocumented submissions must not score. A narrow 5-provider x 4-repair-type launch needs 400-600 verified submissions (13,300-60,000 claim-holder visits at 1-3% contribution); the full 20x8 scope needs 3,200-4,800. Worse, the payer (pre-purchase shopper) and the contributor (mid-dispute claim holder) are different visitors on different searches with no cross-subsidy, and the give-to-get wall that fixes the cold start destroys the consumer revenue line.

Legal exposure and the badge (stress-tested, HIGH). Publishing user-submitted aggregates is defensible but not free: Section 230 is a defence, not a shield from being sued, and a publisher-side defamation defence reaches ~$250,000 by the eve of trial against a $26K-$90K revenue ceiling. Anti-SLAPP fee shifting is unavailable in the 7th Circuit (Endurance, Northbrook IL) and unsettled in the 8th (CarShield parent NRRM, Saint Peters MO, where the old MO statute covered only quasi-judicial proceedings and the new UPEPA is weeks old and untested). Section 230 also stops at the operator's own editorial ranking prose and does not bar FTC action under 16 CFR Part 465 (effective Oct 21 2024, penalties to $51,744 per violation). The paid "verified good-actor badge" must be deleted rather than managed: it is the operator's own commercial speech about a paying customer, triggers material-connection disclosure under 16 CFR 255.5, and reproduces the exact structure that broke the BBB in 2010 (paid accreditation raising letter grades; an A-minus obtained for a fictional business registered under a designated terrorist organisation's name) and drew RICO litigation against ConsumerAffairs' ~$3,000/month accredited-brand programme. Both organisations denied wrongdoing and neither kept its reputation. Removing the badge costs 13-23% of modelled revenue and all recurring revenue, leaving licensing with one named prospect, no established price, and a subpoena route that substitutes for paying once a case is filed.

Biggest killer: legal-cost-to-revenue asymmetry (HIGH). A sub-$100K-revenue publication naming companies with revenue estimated in the hundreds of millions, both already litigating, in forums where anti-SLAPP protection is weak or untested, with exposure that grows as the data gets more accurate.

What would flip the verdict. Two tests runnable for under $15,000 before any build: (1) a signed LOI at $8,000+ from a plaintiff firm or research group for aggregated claim-outcome data; (2) 400+ document-verified submissions across 5 providers and 4 repair categories within nine months via a plaintiff-firm intake partnership. Clear both and this becomes FEASIBLE.

Auto Warranty Claim-Truth Database

Track: Data Product | Market: overseas | Status: PENDING_RESEARCH | Created: 2026-07-22T09:00:00Z | Updated: 2026-07-22T09:00:00Z

One-liner

Discovery Method

Method: Trend Sniffer + Pain-point Extractor + Idea Generator
Signal (Trend): The FTC settled deceptive-advertising charges against CarShield and its claims administrator American Auto Shield for close to $10 million in July 2024, and on December 11, 2025, five weeks before this scan, mailed refund checks totaling more than $9.6 million to 168,179 consumers. That is a live enforcement story, not a stale one, and it is happening in the single largest player in a $25-36 billion US industry.
Signal (Pain): Two independently checked complaint aggregators (pissedconsumer.com at 1.4/5 across 254 reviews, and the BBB customer-review layer reported at 2.13/5 even while the company holds BBB "A+" accreditation) both surface the identical theme in first-person quotes: claims marked "not covered" after the fact, deductibles that only apply "if they approve the claim," and repair shops refusing to deal with the administrator over parts and labor-rate disputes.
Evidence: assets/evidence.md, direct-fetched quotes and figures from search.ftc.gov, carshield.pissedconsumer.com, automoblog.com, and checkbook.org.

Demand Details

Who: US car owners (median vehicle age is now well past the manufacturer warranty window) shopping for a vehicle service contract, plus the much larger pool who already bought one and are trying to figure out, mid-dispute, whether their denial is normal or a pattern for that company.

What they want: to know, before handing over $80-150 a month for three to five years, whether a given provider actually pays out or mostly finds reasons not to. Right now that information does not exist in usable form. It exists as scattered one-star reviews, a BBB page that conflates an accreditation letter grade with a separate and much lower customer-star average, and a string of near-identical "Top 10 Extended Warranty Companies of 2026" articles that take referral fees from the very companies with the worst complaint records.

How they express it: directly, in the reviews themselves. "None of the problems my car had were covered under my car insurance," one reviewer wrote, despite paying $200 a month. Another flagged the deductible bait-and-switch: "they advertise that only 100.00 deductible BUT that's only if they approve the claim, which is not mentioned in their advertisements." A third chased a promised refund for months after cancellation. The pattern repeats enough that the FTC built a federal case on it, and repeats enough that a nonprofit consumer-research group (Consumer Checkbook) gave up trying to rank providers and just tells people not to buy the product category at all, rather than building the comparison tool this opportunity proposes.

The gap: every "comparison" resource found in this category (Car Talk, CoverageX, trustedcompanyreviews, and the dozen or so lookalike sites behind them) is a content-marketing funnel that earns a commission when a reader signs up with the company being reviewed, including the ones with sub-1.5-star aggregate ratings. Consumer Checkbook is the one credible non-affiliate voice in the space, and it does not attempt a structured claims database, just an editorial warning. Nobody has built the neutral, submission-based ledger that would let a shopper filter by provider, repair category, and vehicle type and see what actually happened to people who filed a claim.

Monetization model:

Free: submit a claim outcome, see the aggregate approval-rate range for that provider and repair category
Paid one-time report ($12-19): full breakdown before signing a contract, including denial-reason frequency and median days-to-payout for the specific provider being considered
Data licensing: aggregated, anonymized claim-outcome datasets sold or provided to consumer-protection researchers, state attorneys general, and plaintiff's firms already litigating this exact fact pattern (the FTC case establishes there is an active audience for this)
"Verified good-actor" badge: a placement fee charged only to providers whose submitted-claim approval rate clears a published threshold, positioned as the inverse of the standard pay-to-be-listed affiliate model so the incentive runs toward disclosure rather than against it

7-Dim Triage Scores

Demand Pull 5 / Acquisition Feasibility 3 / Agent Advantage 4 / Low-Volume Economics 4 / Operator Hand Lightness 3 / Market Trend 4 / Policy Redline 4 -> Total 27/35

Score rationale:

Demand Pull 5: A federal enforcement action with an active refund cycle, two independently fetched complaint aggregators both surfacing first-person denial complaints, and a credible nonprofit's editorial capitulation ("just don't buy this") all point the same direction. This is about as strong as scout-stage evidence gets for a consumer pain point.
Acquisition Feasibility 3: The organic-search terms that matter most ("is CarShield legit," "best extended warranty company") are already dominated by affiliate content sites with years of SEO investment, so ranking above them is a real climb. Offsetting channels exist: the FTC news cycle itself, personal-finance and auto-forum communities skeptical of extended warranties, and co-marketing with consumer-law firms already running CarShield-adjacent complaint content, none of which have a reason to protect the incumbent affiliate sites.
Agent Advantage 4: Continuously monitoring dozens of providers' BBB/FTC/complaint-aggregator activity, normalizing submitted claims into structured outcome records, and generating provider scorecards is exactly the kind of always-on, structured-data work an agent pipeline does far more cheaply than a human research team.
Low-Volume Economics 4: Near-zero marginal cost per report once the pipeline exists; a handful of paid $12-19 reports or a single research-licensing deal can cover hosting and moderation costs at low volume, unlike a venture-scale ad-supported content play.
Operator Hand Lightness 3: Needs a human legal read on defamation/UGC-liability posture and disclaimer language before launch (naming specific companies by claim-outcome statistics invites pushback even when factually defensible), but day-to-day operation, moderation, and report generation can run largely on the agent pipeline after that.
Market Trend 4: The vehicle service contract market is growing at a mid-single-digit CAGR, the average US vehicle age keeps climbing past the factory-warranty window (more owners becoming VSC shoppers every year), and the FTC's own enforcement posture keeps the underlying pain point in the news rather than fading.
Policy Redline 4: Not 5, because publishing company-level claim-outcome statistics from user submissions carries a real, if legally defensible, risk of cease-and-desist letters from named providers (the BBB/Trustpilot/Yelp precedent shows this model is legal under Section 230 in the US, but it is not friction-free). No hard redline: this is factual claims benchmarking, not insurance brokerage, financial advice, or legal advice, provided the product stays strictly an aggregator and never brokers or sells a vehicle service contract itself.

Downstream Hints

Key assumption to falsify: will past VSC buyers actually submit real claim outcomes (denial letters, payout amounts) at meaningful volume? This is the same cold-start risk already logged for this library's other crowdsourced rate/outcome databases (lane 6 freelance rates, lane 68 streamer sponsorships). Research should check submission rates on any existing informal complaint-sharing venues (BBB's own "add a complaint" flow, ConsumerAffairs' review-submission volume) as a proxy for willingness to disclose.
Known "competitors" are either affiliate-funded comparison content (Car Talk, CoverageX, trustedcompanyreviews, and similar lookalike sites, all of which earn referral commissions from the companies they rank) or one credible non-affiliate editorial voice with no structured database (Consumer Checkbook, confirmed via direct fetch to lack a claims-comparison tool). None hold a real crowdsourced claim-outcome ledger. Research should verify this holds up under a deeper competitive sweep, since this library has previously found "no competitor" claims at scout stage to be hypotheses that a fuller search occasionally overturns (see this library's Netherlands DBA lane lesson on exactly this failure mode).
Reconcile a data conflict flagged in the evidence: one search-tool-only source cites CarShield with a BBB "F" rating, while a directly-fetched source (automoblog.com) shows CarShield holding BBB "A+" accreditation alongside a separate, much lower 2.13/5 customer-review-star average on the same profile. These are likely two different BBB metrics (accreditation letter grade vs. customer review average), not a contradiction, but research should confirm directly on the live BBB profile page before publishing either figure.
Compliance framing for downstream stages: display aggregated approval-rate ranges and anonymized outcome summaries by default; require documentation (a denial letter, a claim number, or similar) before counting a submission toward a scored aggregate, to blunt fabricated or competitor-planted entries; carry an explicit "user-submitted, not independently audited" disclaimer; never draft or file an insurance/legal claim on a user's behalf, and never take referral revenue from a low-scoring provider in exchange for suppressing its data, since that would recreate the exact conflict of interest this product is positioned against.
Phase-2 expansion signal, not part of this scored opportunity: the identical structural gap ("no federal or public database publishes exact denial percentages") was found to exist in the closely adjacent home warranty vertical (American Home Shield, 2-10 Home Buyers Warranty, First American), a considerably larger and equally complaint-heavy category. Worth a dedicated look once the vehicle service contract version proves out the crowdsourcing assumption.

assets/ Evidence List

assets/evidence.md, direct-fetched quotes and figures from search.ftc.gov (FTC settlement + Dec 2025 refund distribution), carshield.pissedconsumer.com (1.4/5, 254 reviews, direct complaint quotes), automoblog.com (BBB/Trustpilot cross-reference, Reddit sentiment summary), and checkbook.org (confirms no structured claims-comparison tool exists in the category); plus flagged lower-confidence search-tool-synthesis figures (comparative approval rates, BBB complaint volumes, market size, home-warranty phase-2 signal) clearly separated from the directly-fetched claims.