Startup Inside — Fund, Validate and Gate a Venture at Bandeira Labs

A four-round, advanced corporate-venturing simulation set inside Bandeira Labs, the ring-fenced innovation unit of Bandeira Logística S.A., a fictional São Paulo last-mile logistics company (R$2.1 billion revenue, 6,400 staff, ~480,000 parcels a day). Bandeira Labs has launched four ventures in three years and killed three of them late, burning roughly R$14 million on products that scaled before they were validated. The CEO has imposed a lean-startup gate: no venture gets growth capital until it shows validated learning, and every venture has a fixed runway with a forced go/no-go gate. You run the first cohort under the new rule, with a hard R$9 million budget and a 9-month runway (burn ~R$1.0M/month). Playing the Corporate-Venture Lead you (1) choose ONE of three pre-screened ideas — RotaSmart (huge market but a 9–12 month enterprise sales cycle selling to your own competitors), EntregaFlex (the sponsor's safe pick, thin margins, validated on one verbal 'interest'), or DevolveJá (an underserved returns problem with uncertain willingness-to-pay) — matching idea to runway and naming the single riskiest assumption rather than chasing market size or sponsor enthusiasm; (2) scope an MVP and a falsifiable test plan, choosing concierge/fake-door minimalism over a R$4M+ full build that consumes the runway, and setting a R$0.9–1.6M validation-sprint budget; (3) read month-3-to-6 evidence that partially contradicts the plan and make the call — persevere toward a R$25M internal Series-A, pivot the segment while runway remains, or kill cleanly — judged on the ≥35% paid-pilot conversion bar and remaining runway, not on sunk cost or vanity metrics; and (4) pitch the gate to a facilitator-played investment committee, defending whether the runway bought learning or just build. The math rewards disciplined, evidence-led, runway-aware lean-startup play and punishes the five classic errors — sponsor-driven selection, market-size seduction, building before validating, vanity-metric validation, and sunk-cost perseverance or a late kill. Final KPIs track Validated Learning, Customer Traction, Runway remaining and Budget burned against the R$9M cap.