SuppleMint

1. One-liner

AI copilot that helps independent body shops win supplement disputes against insurance underpayment.

2. Trend signal — why now?

The collision repair industry is in a payment crisis. Insurers are deploying AI to lowball estimates, and independent shops — 72,000 of them, representing 68.7% of all US locations — are getting squeezed harder than ever.

• 63% of vehicle repairs require at least one supplement, and each supplement triggers a multi-day back-and-forth with adjusters who are incentivized to deny.
• Short pays are “worse than ever” according to shops nationwide (Autobody News, 2025). Only 29.2% of shops working with State Farm report getting paid for OEM-required safety inspections.
• Customer-pay repairs jumped from 8% to 20% of repair orders among tracked shops — meaning shops are increasingly eating the cost or pushing it to consumers because insurers won’t pay.
• 12-state regulatory pilot launched March 2026 to examine how insurers use AI for claims decisions — California, Florida, Texas among them. Regulators are catching on to insurer AI, which means shops now have leverage if they can document properly.
• AI vision technology now matches experienced estimators at 85-90% accuracy for damage identification from photos (Autobody News, 2025). The tools to fight back exist — they just haven’t been packaged for shops.

Provenance:

• Signal 1: Body shops nationwide report intensified insurer pushback and short pays; customer-pay repairs doubled to 20% — Autobody News — 2025
• Signal 2: AI computer vision for damage assessment reaches 85-90% accuracy, matching experienced estimators; 60%+ of shops expected to use AI by late 2026 — Autobody News / Collision Repair Mag — 2026
• Signal 3: 12-state regulatory pilot examining insurer AI in claims decisions; $76.9B US collision repair market — Autobody News — March 2026 Category: Workflow automation

3. The opportunity

Independent body shops are in an asymmetric fight. Insurers have CCC Intelligent Solutions ($6B+ public company) writing AI that automatically generates lowball estimates. Shops have… a phone and some frustration.

The gap isn’t in estimating — CCC ONE and Mitchell already do that (at enterprise prices). The gap is in supplement negotiation: once the insurer denies or short-pays, the shop owner spends 2-4 hours per repair gathering photos, citing OEM procedures, drafting appeals, and arguing on the phone. Most small shops don’t have a dedicated estimator, so the owner does this themselves.

SuppleMint sits in the gap between “insurer denied your supplement” and “you get paid.” Upload your damage photos, the AI matches them against OEM repair procedures, generates a bulletproof supplement package with citations, photos, measurements, and regulatory references — and gives you talk tracks for the adjuster call.

4. Target market

• Primary customer: Owner-operators of independent collision repair shops in the US, 1-10 employees, $500K-$3M annual revenue, not part of a DRP (Direct Repair Program) network.
• Why they buy: Every denied supplement costs them $600-$1,200 in lost revenue per repair. They spend 2-4 hours per supplement on documentation that still gets denied 30-40% of the time. They’re losing to insurer AI and they know it.
• Rough TAM reasoning: ~72,000 independent shops in the US. If 30% are non-DRP shops actively fighting supplements (the ones who feel the pain most), that’s ~21,600 potential customers. At $199/mo, that’s a $51M addressable market.
• Why now for them: Insurers are tightening payment more than ever post-COVID inflation, repair costs are surging due to ADAS technology in modern vehicles, and the 12-state regulatory pilot gives shops new ammunition if they can document insurer AI decisions properly.

5. Product sketch (MVP)

• Photo-to-supplement pipeline: Upload teardown photos and the AI identifies hidden damage, matches it against OEM repair procedures, and auto-generates a supplement request with citations.
• OEM procedure lookup: Searchable database of manufacturer repair procedures — the AI highlights the ones relevant to your specific damage and vehicle, formatted for insurer submission.
• Supplement document generator: One-click PDF with photos, measurements, OEM references, and itemized cost breakdown — exactly what adjusters need to approve.
• Adjuster talk-track generator: AI-written scripts for the phone call with the adjuster, pre-loaded with the specific OEM citations and regulatory references that counter common denial reasons.
• Denial tracker dashboard: Log every denial with reason codes, track your win rate, identify which insurers deny most and on what procedures.
• Insurer pattern database: Crowdsourced (anonymized) data on which insurers deny which procedures in which states — so shops can pre-empt denials before they happen.

6. AI angle — what’s load-bearing

Without AI, a shop owner manually searches through 50-page OEM repair manuals, cross-references them with insurer denial letters, formats photos into a submission package, and writes the appeal narrative. This takes 2-4 hours per supplement.

With AI, the owner uploads photos + the denial reason, and the system generates a complete, citation-backed supplement package in 5 minutes. The AI is doing the hard work: computer vision to identify damage types from photos, NLP to parse OEM repair procedures and match them to the damage, and document generation to produce insurer-ready packages.

Remove the AI and this product doesn’t exist — it becomes a Dropbox folder with some templates.

7. Localization angle (if any)

N/A — this is a US-only play. The insurer-shop adversarial dynamic, OEM repair procedures, and regulatory framework are specific to the US collision repair industry. International markets have fundamentally different insurance structures.

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

• Pricing: $199/month per shop (Pro tier) or $99/month (Starter — limited supplements/month). Enterprise tier at $399/month adds multi-location management and analytics.
• ACV: ~$2,400/year (Pro) or $1,200/year (Starter).
• Rough math to $1M ARR: 420 Pro shops ($2,400 x 420 = $1.008M) or 550 mixed Pro/Starter.
• Rough math to $5M ARR: ~2,100 Pro shops, or ~1,500 Pro + 1,200 Starter. That’s roughly 10% of the non-DRP independent shop market — ambitious but realistic if the tool demonstrably recovers $1,000+ per month in previously-denied supplements.
• Expansion path: Per-supplement success fee (5% of recovered amount above a baseline), multi-location chains, and eventually a marketplace connecting shops with independent appraisers.

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

• Collision Repair Technicians United Facebook group — the largest shop-owner community on social media. Post case studies showing “before SuppleMint vs. after” supplement approval rates. Shop owners share tools here obsessively. Target the most active posters with early access.
• AutoShopOwner.com forum — direct outreach to shop owners posting about insurer disputes. Offer 30-day free trial with guaranteed ROI: if the tool doesn’t recover at least $500 in one month, it’s free forever.
• Autobody News and BodyShop Business — these are the two trade publications every shop owner reads. Pitch a case study article: “How One Shop Recovered $4,200 in Denied Supplements in 30 Days Using AI.” Trade press in niche industries drives real sales.
• State collision repair associations — there are 50 of these, each with member directories. Offer a “member discount” partnership. Present at their annual conferences.
• Cold outreach to shops with bad Google reviews mentioning insurance delays — these are shops actively losing business because of the supplement problem. Scrape review sites, identify the pain, send a personalized demo.

10. Build complexity — justification

Medium. The core AI pipeline (photo analysis + OEM procedure matching + document generation) uses off-the-shelf vision APIs and LLMs, but the OEM repair procedure database requires significant data ingestion and structuring. The document generation engine needs to output insurer-specific formats. A pair of engineers could ship a credible v1 in 10-12 weeks — the first month is OEM data pipeline, the second is the AI workflow, the third is the shop-facing UI and insurer submission formatting.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Generating documentation and recommendations is legal. Not practicing law or adjusting claims.
Ethical — no harm / dark patterns	✅	Helps shops get fairly paid for work performed. Fights insurer underpayment, not fraud.
Market exists (evidence above)	✅	72K independent shops, $76.9B market, documented underpayment crisis.
1-5 person team can build this	✅	2-3 engineers + 1 industry advisor.
Launchable with <$50K / ₹40L	✅	Cloud infrastructure + API costs. No hardware. Main cost is AI API usage and OEM data licensing.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	17/20	Hair-on-fire for non-DRP shops. Every denied supplement is $600-$1,200 in lost revenue. Owners personally waste hours on documentation. Burnout-level frustration.
Demand evidence	15	12/15	Strong industry press coverage, trade association advocacy, shop owners openly discussing the problem. Customer-pay doubling is a proxy signal. No one’s buying a direct “supplement AI” yet because it doesn’t exist — but the pain is undeniable.
Build feasibility	15	11/15	Off-the-shelf vision and LLM APIs, but OEM procedure database is a meaningful build. Insurer-specific formatting adds complexity. 10-12 weeks for a credible v1 with a pair of engineers.
Distribution clarity	15	12/15	Tight community (Facebook groups, forums, trade press, state associations). Shop owners share tools aggressively. Concrete channels with realistic conversion paths.
Revenue mechanics	15	11/15	$199/mo is well within a shop’s tool budget ($100-300/mo baseline). ROI is directly measurable — if you recover one denied supplement per month, the tool pays for itself 3-6x. But proving attribution requires a few months of data.
Time to first revenue	10	7/10	Revenue in 6-8 weeks post-launch via free trial → paid conversion. Need to demonstrate ROI in trial period. Not instant, but fast for a B2B SaaS.
Defensibility	10	6/10	Soft moat from crowdsourced insurer denial pattern data (gets better with more shops), OEM procedure database, and workflow lock-in. A 6-month head start matters. But CCC or Mitchell could build this as a feature.
Total	100	76/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · sales-heavy

Needs strong AI/ML engineering for the vision + NLP pipeline, and a founder who can sell directly to skeptical shop owners who’ve been burned by software promises before. A co-founder or advisor with collision repair industry experience is nearly essential.

Key assumptions to validate (3-5)

1. Assumption: Non-DRP independent shop owners will pay $199/mo for a supplement negotiation tool. How to test: 50 cold calls to shop owners from the Collision Repair Technicians United Facebook group. Target: 15%+ express “I’d try that.”
2. Assumption: AI-generated supplement packages will meaningfully improve approval rates vs. manual documentation. How to test: Run 20 real supplements through the tool alongside manual submissions, compare approval rates and time-to-payment.
3. Assumption: OEM repair procedure data can be legally ingested and searchable at scale. How to test: Legal review of OEM data licensing terms (I-CAR, AllData, Mitchell) and technical feasibility of building the database in month 1.
4. Assumption: Shop owners will adopt a new tool amid already-stretched workflows. How to test: Observe onboarding friction in first 20 beta users. Target: 80%+ complete first supplement within 48 hours.

Risk flags

1. [Platform risk]: If CCC Intelligent Solutions or Mitchell add a supplement negotiation feature, they have distribution to 100K+ shops instantly. Mitigation: move fast, build the crowdsourced denial database moat, and position as “for independents, against the insurer-industrial complex.”
2. [Data licensing risk]: OEM repair procedure access may require expensive licensing from AllData, I-CAR, or similar. Could cap margins early on. Mitigation: start with the most common vehicle makes (Toyota, Honda, Ford, Chevy — covers ~60% of repairs) and expand.
3. [Regulatory uncertainty]: The 12-state insurer AI pilot could lead to new regulations that either help (more transparency = more ammunition for shops) or hurt (restrictions on AI in claims could apply to both sides). Mitigation: position as a documentation tool, not an AI adjuster.

14. Structured verdict

Score:                  76/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical founder with collision repair industry advisor or co-founder
Time to revenue:        8-10 weeks post-launch
Capital to launch:      $15-25K (AI API costs, cloud infra, OEM data licensing seed)
Top 3 assumptions to validate first:
  1. Shop owners will pay $199/mo — validate with 50 cold calls (target: 15%+ interest)
  2. AI supplement packages outperform manual ones — validate with 20 A/B test supplements
  3. OEM procedure data is legally and technically accessible — validate with legal review + prototype
Kill criteria:
  - Abandon if <10% of 50 cold outreach calls show willingness to try
  - Abandon if AI-generated supplements don't improve approval rate by at least 20% vs. manual
  - Abandon if OEM data licensing costs exceed $3,000/month at launch scale

15. Next step — 1-week validation sprint

• Day 1-2: Join Collision Repair Technicians United on Facebook and AutoShopOwner.com. Post a survey: “How many hours/week do you spend on supplement documentation? How often do insurers deny your supplements?” Message 30 active shop owners directly. Target: 50+ responses, 10+ DM conversations.
• Day 3-4: Build a clickable prototype (Figma) showing the photo-upload → supplement-generation workflow. Show it to 10 shop owners via Zoom. Record their reactions, especially: would they pay $199/mo, and what’s their current process.
• Day 5: Score the results. Go if: (a) 8+ of 10 shop owners say “I’d try this,” (b) average reported supplement time is >2 hours, and (c) average denial rate is >25%. No-go if fewer than 5 of 10 express genuine interest — the pain might be real but the willingness to adopt new software might not be.