ArbPack Clear — Wholesale Auction Arbitration Claim Pack Desk

Done-for-you arbitration claim completeness & evidence pack for independent used-vehicle dealers buying at Manheim, ADESA/OPENLANE, ACV, and related NAAA-policy auctions — so procedural mistakes stop killing valid recoveries after June 1, 2026 guideline changes.

Final decision

BLUEPRINT

Proceed. Clears evidence threshold and six-gate rubric (27/30). Distinct from OEM warranty chargeback desks, title/lien desks, and generic dealer consulting. Service-first, outcome-priced, AI-native documentation desk with human claim-specialist RELEASE.

Executive summary

Independent dealers source a large share of inventory from wholesale auctions. When condition reports omit structural, mechanical, title, or odometer issues, auction arbitration is the contractual recovery path — but claims die on process: missed 10-day digital filing windows, missed 2–5 day evidence milestones, mileage-cap violations, wrong filing channel, vague defect narratives, or repairing before resolution. ArbPack Clear sells a done-for-you Arbitration Claim Pack: eligibility screen against current NAAA + platform rules, evidence checklist completion, claim narrative, portal-ready upload kit, and milestone calendar — reviewed/released by a trained arbitration specialist. The dealer (or their authorized auction account user) remains the claimant; ArbPack never practices law and never represents the dealer before a court.

Thesis

Wholesale auction arbitration is a high-stakes, deadline-driven, document-and-rule synthesis workflow where most value is lost to procedural failure—not to absent defects. AI can extract condition-report language, map defects to NAAA/platform eligibility rules, draft precise claim narratives, and run completeness/QA checks faster than a busy lot manager; a human specialist remains the trust chokepoint for RELEASE. As models improve at multimodal photo/VIN matching and policy retrieval, unit COGS falls while claim quality rises — satisfying the Sam Altman test without becoming a customer-operated co-pilot.

Discovery rationale

This run steered away from the manifest’s heavy regulatory-filing/compliance-engine skew into local/independent dealer back-office recovery operations. Fresh research (12+ queries) across logistics D&D (already saturated in manifest), ASC bill-only implants (duplicate of existing desk), pharmacy DIR (SaaS-crowded), hospital 501(r) FAP (duplicate), HHG mover FMCSA (duplicate), energy-broker commission (UK-skewed evidence), ATM remittance (weak WTP), and wholesale auction arbitration (open + dated June 2026 NAAA trigger). Auction arbitration won on: active buyer process pain documented by trade guides; existing spend on arbitration fees/labor; narrow MVP; remote fulfillment; outcome pricing; and zero Manheim/ADESA/NAAA semantic matches in the 593-run manifest.

Candidate comparison

Scoring used the 20-factor 1–5 rubric (max 100). ArbPack led on novelty, MVP clarity, distribution, active demand, and Altman/test fit.

CODE validation

• Consumer/buyer trend: Digital wholesale buying (Simulcast/OVE/ACV) expands remote purchases where defects surface after transport — arbitration windows collide with logistics delays (Ready+2 exists because this pain is structural).
• Opportunity: Dealers lose recoveries on process (deadlines, evidence, narrative precision, mileage caps), not only on merit. June 2026 NAAA changes raise the bar for high-value defects and disclosure quality.
• Demand: Manheim/ADESA publish detailed arbitration guides; AutoAuctionAtlas documents “seven mistakes that kill claims”; ADESA charges ~$195 invalid-claim fees; dealers already pay platform arbitration fees and burn GM/lot-manager time.
• Economic sizing (range, uncertain): ~52k active independents; if 15–25% buy regularly at major digital auctions and 5–15% of purchases trigger an arbitration review need annually, serviceable units are tens to low hundreds of thousands of claim-events/year. At $500–$1,200/pack with 10–20% success-fee optional add-on on recovered dollars (commercial services, not legal contingency), even a 0.5–2% share supports a meaningful multi-million ARR path. Figures are order-of-magnitude; conversion unknown until pilots.

Rubric scorecard (six gates)

Total: 27/30. Anti-commoditization: generic ChatGPT cannot hold portal credentials workflows, milestone calendars, platform-specific invalid-fee avoidance, or accountable RELEASE under a services MSA.

Target buyer

ICP: U.S. independent (non-franchise) used-vehicle dealers, 8–120 retail units/month, buying ≥4 vehicles/month via Manheim digital, ADESA Clear/OPENLANE, and/or ACV; 1–3 locations; no dedicated arbitration clerk.

Economic buyer: Dealer principal / GM (owns floorplan risk and inventory P&L).

Champion: Inventory manager, buyer, or office manager who receives the car and discovers the defect.

Beachhead: Sun Belt independents (TX/FL/GA/CA) buying digitally with third-party transport — highest deadline collision rate.

Jobs-to-be-Done

• When an auction unit arrives with an undisclosed defect, help me file a complete, on-time, platform-correct arbitration claim without burning my buyer’s day.
• When NAAA/platform rules change, keep my claims eligible without me reading policy PDFs.
• When the arbitrator asks for more evidence in 2–5 days, respond completely the first time.
• Avoid invalid-claim fees and “Buyer Bought” closures from silence/missed milestones.

Painful problem

Valid arbitration rights expire on a calendar, not on fairness. Digital windows are typically 10 calendar days (sale day = day 1); evidence milestones can be 2 business days (ADESA) or 5 calendar days per Manheim milestone; mileage caps (e.g., ~100–150 miles on some platforms) and “do not repair” rules void claims. Guides explicitly list missed deadlines, wrong channels, vague descriptions, and premature repairs among the top claim killers. June 1, 2026 NAAA updates add a 2% of purchase price per-defect threshold for $50k+ vehicles and prohibit generic “No Arb for X” seller announcements — raising both buyer opportunity and drafting precision requirements.

The outcome we sell

A specialist-released, portal-ready Arbitration Claim Pack for a specific VIN/stock number: eligibility determination vs current rules, defect-to-rule mapping, evidence completeness score, draft claim narrative, upload checklist, milestone calendar, and recommended resolution ask (return / price adjustment / seller-funded repair at wholesale rates). Optional: milestone response packs during the open case. Success metric = pack accepted for filing + arbitrator progression without procedural denial; secondary = $ recovered / sale unwound (dealer-reported).

First one-feature MVP wedge

Evidence summary

• Verified: NAAA/Manheim June 1, 2026 guideline changes; platform filing/evidence windows; ADESA invalid claim fee ~$195; 7M+ 2025 AuctionNet sales; ~52–53k active independents / 9.8M independent retail sales 2025; independents ~42% of auction buyers (2023 Gitnux citation) and ~65% of ACV buying activity (mid-2025 secondary summary).
• Inferred: Material share of lost recoveries are procedural; beachhead willingness to pay $350–$1,200 when repair estimates exceed $800–$2,000+.
• Unverified: Industry-wide arbitration filing/denial rates; exact % of digital purchases that become claims; incumbent third-party DFY pricing (no clear national DFY incumbent found).

Claim table

Source-claim matrix

Market and demand evidence

Wholesale auctions remain the primary acquisition channel for many independents. AuctionNet reported just over 7 million U.S. wholesale units in 2025 (highest since 2019). Independents sold 9.8 million retail units in 2025 with ~52–53k active dealers. Digital channels expand remote defect discovery after transport — structurally colliding with short arbitration clocks. Policy complexity increased on June 1, 2026. Demand shows up as: detailed platform help centers; paid invalid-claim fees; dealer education content enumerating procedural failures; and inventory-manager labor already allocated to claim paperwork.

Active buyer conversations

• Platform help centers and buyer/seller arbitration guides (Manheim, ADESA) — continuous operational Q&A.
• Trade explainers (AutoAuctionAtlas) walking dealers through claim-killer mistakes and multi-platform deadline tables.
• NIADA/state independent dealer association education tracks on wholesale buying risk.
• Dealer forums/Facebook groups commonly discuss “missed the window” / transport delay / condition-report disputes (qualitative; treat as directional, not quantified).

Competitive landscape

• Auction arbitration departments (Manheim/ADESA/ACV) — process owners, not dealer-side DFY pack assemblers.
• Guides & checklists (AutoAuctionAtlas, blog content) — education, not fulfillment.
• Franchise attorneys / dealer counsel — engage on large OEM chargebacks or litigation; overkill/slow for routine auction arb packs.
• Internal GM/buyer labor — default incumbent; high opportunity cost and error rate under deadline pressure.
• Generic AI chat — can draft text but cannot own evidence QA, deadline orchestration, or RELEASE accountability.

Competitor and budget validation

Budget already exists as: (1) platform arbitration fees; (2) invalid-claim fees (~$195 ADESA); (3) third-party inspections/estimates; (4) transport costs at risk; (5) GM/buyer hours; (6) occasional attorney time on large units. ArbPack redirects a slice of (5) and prevents loss of vehicle equity/(1)–(4). Win condition: faster, more complete packs than DIY under the new HV thresholds, with measurable reduction in procedural denials — not “no competitors.”

Pricing evidence and proposed pricing

Never hourly. Price anchors to avoided loss (typical mechanical threshold $800; HV defects often $1k–$5k+ repair; full unwind can be entire hammer price). ADESA invalid fee and inspection costs prove buyers already pay for process mistakes.

Regulatory and compliance considerations

Core regime is contractual auction policy (NAAA Arbitration Guidelines + platform addenda), not state UPL for preparing dealer’s own claim materials. Still: avoid legal advice; avoid representing the dealer in court or administrative dealer-board OEM protests (different product — franchise law). Respect platform ToS on account access (prefer dealer-uploaded evidence + dealer submits, or limited delegated access with written auth). Handle dealership data (VIN, buyer account IDs) under a commercial MSA + DPA-lite. Do not fabricate evidence. Do not coach fraudulent claims.

Licensing boundary

Disclaimers: ArbPack provides operational documentation support for platform dispute processes. Not a law firm. Not a guarantee of claim approval or recovery. Dealer remains solely responsible for portal attestations and vehicle custody rules.

AI-native advantage

AI changes economics by compressing the “60-minute documentation sprint” and multi-policy synthesis that currently requires scarce buyer attention. Multimodal models compare defect photos to CR language; retrieval keeps NAAA/platform matrices current after June 2026 changes; deterministic rules encode deadlines/mileage caps; the specialist only touches exceptions and RELEASE. Throughput per specialist rises nonlinearly vs pure labor shops.

Internal AI engine architecture (10 layers)

1. Intake — secure upload (CR PDF, photos, sale confirm, transport BOL, estimate).
2. Normalization — VIN validate, stock #, sale date, platform, sale light, odometer.
3. Retrieval/knowledge — versioned NAAA + Manheim/ADESA/ACV rule cards; threshold tables.
4. AI workbench — defect tagging, CR contradiction spotting, narrative draft, evidence gap list.
5. Deterministic rules — filing clock, evidence clock, mileage cap, HV 2% calc, as-is/red-light exclusions.
6. Human chokepoint — specialist RELEASE / rewrite / decline.
7. QA — checklist score ≥ threshold; red-team “claim-killer” scan.
8. Delivery — Pack PDF + portal field map + zip + calendar ICS.
9. Learning loop — outcomes tagged (procedural denial / merit denial / adjustment / unwind) → prompt/SOP updates.
10. Model-portability — provider-agnostic LLM/vision adapters; rules DB separate from model weights.

AI-vs-human operations pipeline

Dynasty translation layer

• Buyer: Independent dealer GM paying to protect auction inventory dollars.
• Service: DFY claim pack + optional desk; automated extraction + human RELEASE.
• Workflow: Intake → rules → draft → RELEASE → dealer files → milestone support → outcome learn.
• Tooling: Day-1: Drive/Dropbox, Notion/Airtable, Slack/email, Claude/GPT vision, calendar; later: thin portal.
• Sales: “We build the arbitration pack before your 10-day clock dies.”
• Delivery: Manual specialist ops first; automate extraction/QA next.
• Expansion: Seller defense packs; multi-auction always-on desk; franchise groups.

Anti-duplication analysis

Checked full manifest.json (593 runs) and repo HTML filenames. Zero hits on Manheim/ADESA/NAAA/auction-arbitration claim packs. Adjacent but distinct: franchised-dealer-factory-warranty-claim-completeness-desk (OEM warranty, not auction arb); dealer-floorplan-trust-compliance-desk; title/lien desks; dealer-out-of-state-title-registration-desk. ArbPack’s buyer+workflow+outcome (auction contractual arbitration pack) does not match those.

Anti-commoditization analysis

If frontier models let dealers self-draft narratives, ArbPack still wins on: continuously updated multi-platform rule matrix; deadline orchestration; evidence completeness scoring; invalid-fee avoidance; outcome-tagged learning; and accountable specialist RELEASE under an MSA. The product is the operating system for claim units, not a chat window.

Service delivery workflow

1. Dealer triggers via form/SMS after defect found.
2. Upload pack within hours; clock starts from sale date (system computes remaining hours).
3. AI eligibility + gap list same day.
4. If gaps: structured ask (photo angles, VIN shot, estimate line items).
5. Specialist RELEASE within SLA (e.g., 6 business hours if ≥48h remain on clock; rush SKU if <48h).
6. Dealer submits; ArbPack watches milestones; Response Sprints as needed.
7. Outcome captured; postmortem if procedural fail.

Operations as product

SOPs per platform; required evidence lists; automated completeness checks; exception queues (as-is/red light, mileage exceeded, HV threshold fail); reviewer assignment by platform; confidence scores; audit trails; gold-standard packs; red-team claim-killer checks; customer-ready templates; RCA on failed units.

No-holes quality engine

• Hard stop if sale light/as-is excludes category.
• Hard stop if mileage cap exceeded (unless platform exception documented).
• Hard stop if repair already performed.
• HV 2% math verified for ≥$50k units post-6/1/2026.
• Narrative must cite CR contradiction with dated screenshot reference.
• Two-person rule on packs >$10k exposure (specialist + senior).

What the human expert actually does

Minimum viable offer

“48-Hour Arb Claim Pack” for one digital Manheim or ADESA Clear purchase: eligibility screen, evidence kit, narrative, portal map, calendar. Founding price $450 flat. Money-back if we miss our RELEASE SLA (not if auction denies on merit). Free Eligibility Scan as lead magnet.

Fulfillment process (first 3 customers)

1. Manual Airtable pipeline + Google Drive folder per VIN.
2. Founder/specialist runs AI extraction in ChatGPT/Claude with vision; pastes into pack template.
3. Spreadsheet rule calculator for clocks/thresholds.
4. Email/SMS milestones via calendar reminders.
5. No custom software. Automate only after 5 packs with timed COGS.

Tools and systems

Day 1: Google Workspace, Airtable, Stripe, Calendly, Slack, Loom, OCR/LLM APIs, Adobe/Preview for PDF assembly, platform policy PDF library. Later: customer portal + webhook sale ingest (optional dealer CSV).

Human-in-the-loop quality control

No pack ships without specialist RELEASE. Auto-decline codes for fatal disqualifiers. Weekly calibration on anonymized gold packs. Customer cannot “force” filing of packs we mark ineligible (protects invalid-fee risk and reputation).

Nonlinear scaling and unit economics

COGS stack: model inference, storage, specialist minutes, QA sample, payment fees, rework, support. Path to 50%+ gross margin is credible by day 90 via template reuse + rules automation. Revenue per FTE target year-1: $250k–$400k with mixed packs/retainers.

Distribution proof table

Sales and outreach plan

Lead with diagnosis, not demo: free Eligibility Scan → 15-min review → single Pack offer if clock allows → retainer after 2 successful packs. Outbound personalization uses public auction activity signals only where legitimate; otherwise association lists and partner intros.

Founder-led content plan

Teach: June 2026 NAAA changes; platform deadline matrices; claim-killer teardowns; HV 2% calculator; Ready+2 myths; photo protocols; “as-is vs green light” mistakes. No generic AI hype.

First 30 days of content

1. 10 posts: NAAA 2% HV math examples; No-Arb announcement ban explained; Manheim vs ADESA clocks; mileage caps; photo protocol; transport delay playbook; vague narrative before/after; invalid-fee avoidance; seller disclosure upgrades; retainer ROI story (hypothetical labeled).
2. 3 diagnostic teardowns: anonymized CR contradiction; missed evidence milestone; HV threshold fail.
3. 2 lead magnets: Claim-Killer Checklist PDF; Arbitration Deadline Calculator (sheet).
4. 1 live review webinar: “June 2026 NAAA changes for independents.”
5. 1 outbound diagnosis template: VIN-agnostic clock risk memo for digital buyers.

Lead magnet and waitlist plan

Magnet: Free Arb Eligibility Scan + Claim-Killer Checklist. Buyer receives a scored gap list and hours-remaining clock before paying. Waitlist CTA for Post-Sale Arb Desk. Scan captures platform, monthly auction volume, and recent miss stories (pain signal). Sales-ready = active clock <7 days or ≥8 auction buys/month.

Warm GTM plan

Convert checklist downloaders; ask recon/transport partners for warm intros; offer founding pricing to first 10 dealers; host association lunch-and-learn.

Targeted outbound plan

50 personalized notes/week to TX/FL/GA independent GMs: reference June 2026 rule change + offer scan. No spray demos. Follow-up = Loom of sample pack (synthetic VIN).

Answer-engine / search visibility plan

Pages targeting: “Manheim arbitration deadline,” “ADESA Clear arbitration evidence,” “NAAA arbitration 2% threshold 2026,” “auction arbitration claim checklist.” Structured FAQ; cite primary policies; updated when NAAA revises.

Pilot design and early-demand-trap mitigation

• Pilot cap: 10 dealers / 25 packs before expansion.
• Incentive: founding pack price; not unlimited custom work.
• Measure: procedural denial rate, specialist minutes, Scan→Pack %, SLA hit rate.
• Refuse custom “fight my seller in court” scope.
• Do not hire bodies to cover broken SOPs — pause at caps.

Early-access feedback flywheel

Every pack outcome tagged. Procedural failures → new QA rules within 72h. Merit denials → eligibility model tuning. Dealer UX friction → intake checklist changes. Product feedback vs custom work gated by change-control board (founder + specialist).

Build-before-scale checkpoints

• After 5 pilots: harden intake + evidence requirements + QA.
• After 10 pilots: harden SOPs, exception queues, reviewer checklists, delivery templates.
• After 20 pilots: pause new logos until COGS, rework, escalation, cycle time measured and within targets.

7-day / 30-day / 90-day launch plans

7-day: Rule library v1 (Manheim+ADESA); pack template; Airtable; 20 outbound; publish checklist; 3 scans.

30-day: 8 paid packs; first partner referral; webinar; ACV rules v1; measure minutes/pack.

90-day: 40 packs; 6 retainers; day-90 COGS ≤$140; procedural denial <10%; decide portal build vs stay manual.

Metrics and KPIs

Scans, Scan→Pack, RELEASE SLA, specialist min/pack, procedural denial %, merit win/adjustment rate (dealer-reported), rework %, gross margin, retainer churn, referral %, NPS, invalid-fee incidents (target 0 caused by us).

Risks and mitigations

Platform ToS/account access risk → dealer submits. UPL perception → disclaimers + decline legal matters. Adverse selection (hopeless claims) → paid eligibility honesty + decline codes. Seasonality of wholesale → retainers. Auction improves UX → still sell orchestration+RELEASE. Contingency-fee optics → optional, capped, commercial framing or drop.

Exhaustive risk register

What could kill this

• Scan→Pack <8% after 80 scans.
• Procedural denial ≥20% after 25 filed packs.
• Specialist >60 min/pack after hardening with no automation path.
• Platforms prohibit third-party preparation in ToS without workable dealer-submit model.
• Cannot acquire dealers under $300 CAC after 90 days.

Go/no-go reasoning

GO. Clear buyer, painful deadline-driven problem, verified policy trigger dated June 1, 2026, existing budget/labor, open vs 593-run manifest, narrow MVP, remote DFY delivery, credible 50%+ margin path, licensing boundary manageable if scoped as documentation support (not legal representation).

Final recommendation

Launch ArbPack Clear as a service-first Arbitration Claim Pack desk for independent digital auction buyers. Start Manheim+ADESA only, founding packs, 10-dealer pilot cap, Free Eligibility Scan magnet, and ruthless procedural-quality metrics before retainers at scale.

Source list

1. Manheim — Latest NAAA Updates (effective June 1, 2026)
2. NAAA — Summary of Changes to Arbitration Guidelines (June 1, 2026)
3. NAAA — Arbitration Guidelines PDF (June 1, 2026)
4. Manheim — Arbitration Process Timelines/Deadlines
5. Manheim — Buyer Arbitration Guide
6. Manheim — Ready+2
7. ADESA — Clear Arbitration Claims
8. ADESA — Arbitration Overview
9. AutoAuctionAtlas — How to File an Arbitration Claim
10. AutoAuctionAtlas — Arbitration Claims Guide
11. Automotive Fleet — AuctionNet 2025 sales
12. NIADA — Independent used sales 9.8M in 2025
13. UseFlai — U.S. dealer counts 2026 summary
14. Blue & Co. — Role of Independent Dealers / NIADA 2025
15. Gitnux — Car Auction Industry Statistics
16. Rework — Used Vehicle Acquisition 2026 Guide
17. Cox Automotive — Q2 2026 Manheim Used Vehicle Value Index
18. MatrixBCG — ACV Auctions target market notes