Operations is where the leverage thesis is proven or broken. The design goal: GMV per ops FTE climbs from $15M to $50M while fill rate and OTIF never dip.
Totals match the business roadmap (~10 → ~150). Ops headcount is the denominator of the leverage metric — it is planned, not discovered.
| FUNCTION | Y1 | Y2 | Y3 | Y4 | Y5 |
|---|---|---|---|---|---|
| Engineering & AI | 4 | 8 | 15 | 26 | 40 |
| Operations (review · sourcing · logistics) | 3 | 6 | 12 | 24 | 40 |
| Sales & marketing | 2* | 5 | 12 | 28 | 50 |
| G&A (finance · legal · people) | 1 | 3 | 6 | 12 | 20 |
| Total | 10 | 22 | 45 | 90 | 150 |
| GMV per ops FTE | $15M | $28M | $38M | $42M | $50M |
*Y1 selling is founder-led; founders counted here.
flowchart TB P["Prospect factory
capability intake"] --> C["Certification verification
scope · expiry · evidence-linked"] C --> Q["Sample & QC review
spec conformance"] Q --> T["Trial PO
small-lot, inspected"] T -->|"pre-shipment inspection pass"| A["FACTORY_CERTIFICATION_APPROVED
on the ledger"] A --> S["Scored continuously
quality · OTIF · cost history"] S -.->|"score decay or expiry"| C
Auto-matching defaults on only when a family has three approved factories — substitution needs somewhere to go.
Forwarder partnerships and 3PL cross-dock; no owned warehouses. Bonded consolidation added in Y4 when lanes justify it.
Pre-shipment sampling on trial and high-risk POs; disputes acknowledged in 24h, resolved with evidence in ≤14 days.
| Y1 | Y3 | Y5 | |
|---|---|---|---|
| Facts auto-approved | 40% | 75% | 90% |
| Reviewer throughput (facts/day, tooled) | 250 | 400 | 600 |
| Review touches per $1M GMV | ~900 | ~350 | ~120 |
| Escalation: material facts | always human — certs, credit, substitutions, money (per HITL policy) | ||