An asset-light stock velocity OS for South African car dealers.

1. The Carvana model and why it does not transplant

Carvana's 2025 financials are remarkable:

• $20.3 billion revenue
• $1.9 billion net income
• 596,641 retail units
• 4.7% net margin (2× industry average)
• Recovery from near-bankruptcy in 2022 to most profitable used-car retailer in US history by 2025

The model has three pillars:

1. Standardised reconditioning. Every Carvana car through a 150-point inspection, reconditioned to Carvana's published standard.
2. Finance attached at browse step. Carvana originates its own auto loans, packages them into ABS, earns finance income.
3. Logistics orchestration. Carvana operates its own delivery fleet, vending machines, inspection centres.

This is a beautiful business when it works. It has also nearly killed everyone who tried to copy it:

• Cazoo (UK) — $8B SPAC 2021, administration May 2024, £260M debt
• Vroom (US) — exited used-car retail January 2024 after years of losses
• Kavak (Mexico) — $8.7B 2021, $2.2B March 2025 (-75%), exited Peru and Colombia
• Shift Technologies (US) — bankrupt 2023

WeBuyCars works becauseit is owned by SA's largest dealer conglomerate with deep capital reserves, two decades of refinement, and physical advantages (60+ buying branches) no startup could replicate.

2. The asset-light alternative

Visio Trust attacks the same problem from the opposite direction. Instead of owning the cars, the warehouses, the loans, and the trucks, we own the coordination layer above the existing SA used-car ecosystem.

1. Verified condition — Visio Inspect
2. Pre-approved finance — Visio Approve + Visio Open Finance
3. Escrow-held payment — Stitch EscrowPay or attorney trust
4. Delivery orchestration — Loop / Picup / WeBuyCars-as-carrier
5. 7-day return guarantee — via escrow state machine

What Visio Trust does not do:

• We do not own a single car
• We do not extend a single Rand of credit
• We do not operate a single delivery vehicle
• We do not store a single vehicle in inventory
• We do not take title to anything

We are the trust + intent + finance + verification + transactionlayer that makes the existing R160B SA used-car market behave like Carvana behaves in the US. The dealer keeps owning their stock. The bank keeps owning the credit risk. The logistics partner keeps owning the trucks. We own the coordination — and the data flywheel that comes with it.

3. The stock velocity insight

The economic case for Visio Trust to a SA car dealer is not the consumer-trust experience. It is stock velocity. Per 2025 AutoTrader Industry Report:

• Appropriately priced SUVs sell within 33-35 days
• Appropriately priced compact hatches sell within 28-30 days
• Typical unoptimised SA dealer holds stock at ~47 days average

Every extra day on the lot is floorplan interest the dealer pays to a bank. At a typical floorplan rate of 12.75% annual on a R416,082 average vehicle:

floorplan_interest_per_day = R416,082 × (0.1275 / 365) = R145.40 per car per day

A dealer doing 40 cars per month who can shave 18 days off their average days-on-lot saves:

40 cars × 18 days × R145.40/day × 12 months = R1,254,490 per year

R1.25M of pure margin recovered for the cost of a R8,000/month Velocity subscription = R96,000/year. The ROI is approximately 13×before you count any other benefits (improved F&I attach rates, fewer fall-through deals, better lead quality from BDC routing, cheaper marketing because Intent tells you when you're above market).

This is the entire sales pitch. It is research-grounded, math-obvious, and aligned with what dealers already care about.

4. The unified transaction ledger

Visio Trust's central technical artefact is the vt_transactionstable — the single source of truth for any car transaction flowing through the Visio Auto suite:

vt_transactions:
  id                          uuid PK
  dealer_id                   → vt_dealers.id
  vin                         text
  vehicle_description         text
  price_zar                   numeric
  buyer_email                 text

  -- Foreign refs into the 5 sibling products
  inspect_report_id           text
  approve_simulation_id       text
  open_finance_assessment_id  text
  bdc_lead_id                 uuid
  intent_price_check          jsonb

  -- Visio Trust's own state machine
  escrow_account_id           text
  escrow_status               text   -- pending|held|released|refunded|disputed
  delivery_status             text
  return_deadline             timestamptz
  status                      text   -- pending|held|delivered|returnable|completed|disputed

Every other product writes its primary key into this table. This is the layer that makes “Visio Auto Suite” a coherent product line rather than five separate apps. A dealer's history of stock velocity, F&I attach rates, customer touchpoints, and escrow throughput all live in vt_transactionsjoined to the sibling tables. Leaving the suite means losing the entire historical view — which is the central switching cost.

5. The escrow architecture

Escrow is the most legally sensitive component. We deliberately do not hold buyer funds ourselves. Instead, funds flow into a regulated third-party escrow facility under one of three patterns:

1. Stitch EscrowPay (preferred). Registered SA payments fintech with an EscrowPay product designed for marketplace use cases. API-driven, programmatic, full state machine.
2. Attorney trust account (backup). Traditional, well-understood, irrefutably legal in SA. Slower and more manual but always available.
3. Direct bank holding account (cheapest). Ozow + dedicated holding account with strict reconciliation. Slight legal ambiguity until tested.

Visio Trust ships with the Stitch path as primary, attorney trust as backup, and a stub mode that flags every escrow event with is_stub: true until the Stitch partnership is signed. The product is fully runnable today and transitions to live escrow without code changes.

6. The 7-day return mechanism

T0:         Buyer pays into escrow → status = pending
T0:         Visio Trust opens escrow account → status = held
T+0 to T+5: Delivery scheduled and executed
T+5:        Vehicle delivered → return_deadline = delivered_at + 7 days
T+5 to T+12: Buyer can request return at no cost
T+12:       Auto-release to dealer minus 1% GMV fee → status = completed

If the buyer requests a return inside the 7-day window, the escrow refunds and the carrier picks up the vehicle. The dealer never had the money in their account, so the refund is friction-free. This is the structural advantage of escrow over post-payment-with-refund: there is nothing to claw back.

7. Pricing and revenue

Target: 500 dealers × blended R8K/month = R48M ARR from subscriptions, plus 1% GMV on 5% of the R160B SA used market = R80M ARR from transaction fees. Combined: ~R128M ARR.

8. Three structural moats

1. Network effects within the suite. Each product makes the others more valuable. The full suite is structurally more valuable than any individual product.
2. The unified ledger as switching cost. A dealer with six months of stock velocity + F&I + touchpoint + escrow history in vt_transactions cannot leave without losing the entire view.
3. Data flywheel feeding Visio Intent. Every Visio Trust transaction adds k-anonymised signal data to the Intent index. The more dealers use Trust, the more valuable Intent becomes to OEMs and banks. Competitors cannot replicate without first building the upstream signal capture across hundreds of dealers — which takes years.

9. What can go wrong

• Stitch EscrowPay partnership delays. Mitigation: ship with stub mode + attorney trust as backup.
• WeBuyCars retaliates. Vertically integrated; building dealer-facing SaaS would cannibalise their own funnel. Expect them to ignore us until scale.
• A major dealer group builds in-house. Motus or Super Group could theoretically build their own. They probably won't because (a) they prefer specialists and (b) network effects favour a neutral cross-dealer layer.
• Honesty Protocol violation. The existential risk. A single fabricated stock velocity number, a single fake testimonial, a single inflated revenue claim — and the entire suite's credibility collapses. Mitigation: demo flags everywhere, source-cited benchmarks, explicit is_stub: true markers in the database itself.

10. Conclusion

Visio Trust is the asset-light answer to a question Carvana, Cazoo, Vroom and Kavak all answered the wrong way. The Carvana experience is the right experience. The Carvana balance sheet is the wrong balance sheet. By owning only the coordination layer — escrow, inspection-as-a-trust-signal, finance pre-approval, delivery orchestration, and a 7-day return guarantee — Visio Trust delivers the Carvana buyer experience to the South African used-car market without any of the capital intensity that has killed every other attempt.

The R128M ARR target is the floor; the ceiling is whatever fraction of the R160B SA used-car market we can convince dealers to route through us. The economic case to the dealer is stock velocity, not consumer trust. The economic case to the buyer is Carvana-grade purchase experience on any car. The legal case to regulators is that we are a coordination layer, not a credit provider, not a bureau, and not a vehicle owner. All three cases are structurally correct.

This is the business.

References: Carvana Q4 and full-year 2025 earnings (investors.carvana.com, 2026-02-18); Cazoo bankruptcy (Fortune, May 2024); Kavak valuation collapse (Bloomberg, April 2025); AutoTrader 2025 Industry Report (stock velocity benchmarks); TransUnion Q4 2025 Mobility Insights; Mordor Intelligence SA Used Car Market 2025 (R160B annual market); Stitch EscrowPay product documentation. Repository: github.com/Iamhamptom/visio-trust, commit 3ddb300. Live deploy: visio-trust.vercel.app.

Cite as: Hampton, D. (2026). “Visio Trust: An Asset-Light Stock Velocity Operating System for South African Car Dealers”. Visio Research Labs, VRL-AUTO-009.