An AI investor-assistant reviews Suprance's own business plan. Harsh-but-fair. No cheerleading. Red flags, yellow flags, green flags — tracked over every revision. Current verdict: HOLD. Here's what it'd take to reach green light.
The plan is grounded: projections benchmarked against real marketplace comps (Toptal, Upwork, Fiverr), moat section split into what's defensible vs what's a feature, Tier 2 review costs modeled into margin, pricing narrowed to $2K–$8K flagship, EU-first commitment. Two things we can't fix in a document: no paying customers yet (pre-launch), solo founder (founding engineer planned, not a co-founder). The condition: close first 10 paying EU projects within 60 days of funding.
Methodology: The investor lens reads the full business plan like a seed-stage partner would: challenge the premise, pressure-test unit economics, flag key-person risk, assess moat durability, triangulate market claims. Scoring uses the same dimensions as a Sequoia / YC intake memo. Output is structured: red (deal-breaker), yellow (monitor), green (resolved or strength). Flags evolve across revisions — we track the journey, not just the snapshot.
Each flag has a category, a 1-line reason, and — if applicable — a note on how it's been mitigated or what would resolve it.
Classic marketplace problem: once a client loves their developer, why stay on the platform? Plan now has a 5-point anti-disintermediation strategy (escrow, scans, retainers, reputation, insurance). But it's still unproven at volume.
Stays yellow until retainer attach rate hits 30%+ over 20 projects.
$250K for 13 months = $20K/mo burn. One missed quarter compresses runway. EU-first simplifies (no dual-compliance), but a larger raise ($350K at same terms) would provide genuine slack.
Monitor cash position monthly. Bridge round if needed at Month 9. Sensitivity table shows stall-case path.
Plan now includes a conservative case (halved growth, 30% higher CAC) that still produces $9.5M Year 3 revenue. But even the conservative case is a model — real data required from first 20 projects.
Stays yellow until 20 projects generate real CAC/LTV data. Flips green at 3×+ LTV/CAC measured.
suprance.com has a live AI scoping engine. That's the single most compelling pre-seed signal: the founder can ship. Many pitch decks at this stage are PDFs; Suprance's is a running application.
Glebs spent months delivering AI-built projects and watching clients hit the 80% wall. The story is specific, personal, believable. This isn't a category pick from a VC thesis deck.
80% of marketplaces fail at cold-start. Suprance's founder network into European SMBs is a genuine structural advantage most pre-seed marketplaces don't have.
Poland, Romania, Bulgaria (intra-EU, no GDPR friction) and LATAM (US timezone) offer €20–35/hr junior devs. Cost arbitrage is structural, not a growth-hack.
Avoiding marketplace structure until Stage 3 saves $25K in MTL, platform ToS, SOC 2. Shows founder understands the Airbnb/DoorDash playbook. Rare at pre-seed.
Network effects, accountability primitive ("signed by"), data moat (spec→outcome loop), vetted supply, category ownership. Multiple layers — if one fails, others still work.
Natural acquirers: Lovable (human layer), Cursor (services attach), Upwork (AI modernization), Toptal (AI pivot), Stripe (Atlas attach). 5+ plausible paths in $100M–$1B range.
Suprance ran its own plan through gstack CEO review and published the 7.5/10 verdict with all failure modes. Few founders document adversarial reviews of their own thesis. Trust signal.
Month 1 hire: senior engineer at $6K/mo + 2% equity, 4-year vest, 6-month cliff. Expected to grow into CTO post-seed. Derisks key-person flag — two people own the company from day 1.
Plan now explicitly commits to EU-only for first 50 projects. US expansion begins in Q3 only after EU repeat-rate exceeds 30%. Eliminates the dual-continent overextension risk.
Working platform live. 5 warm-intro pilot projects in pipeline. 6 adversarial research memos published. Junior supply pipeline identified across EU + LATAM. Pre-revenue ≠ pre-proof.
LLM layer abstracted behind provider-agnostic interface. Switching from Claude to OpenAI/Google = 1-day code change. Post-seed: add A/B model testing + automatic failover.
One product in Year 1. Fix-tier ($99-499) is an acquisition wedge only. Enterprise ($25K+) explicitly deferred to Year 2. One sales motion, one delivery pipeline, one margin model.
Added conservative scenario: halved growth + 30% higher CAC → still produces $9.5M Year 3 revenue at 40% margin. LTV/CAC 10× at Year 3 (vs. 33× base). The business works even pessimistically.
The investor lens tracks changes. Red → Yellow → Green isn't guaranteed — some issues remain, some get worse. Here's the honest journey.
First version of Suprance. Strong thesis but nothing built yet.
Running the plan through the gstack CEO review framework publicly. Identified 6 kill zones honestly. No cheerleading.
Platform shipped. Executive summary written. 3-year P&L, cash flow, sensitivity analysis. Detailed hiring sequence with triggers.
Founder added a "concierge escape hatch" — deferring MTL, platform ToS, SOC 2 until Stage 3. Saved $25K of legal, redirected to marketing.
Founder revealed structural advantages not previously in plan: warm access to European SMBs + easy access to junior devs at €20–35/hr. Moved strategy to dual-market (EU anchor + US scale).
Major revision. Projections benchmarked against real marketplace comps. Moat split into "actually defensible" vs "features." Tier 2 review cost modeled into margin. Pricing narrowed to $2K-$8K flagship. EU-first. Founding engineer on CTO track.
The investor's specific asks before flipping HOLD → INVEST. These are the deltas that matter.
This remains the single most important milestone. Close first 20 EU projects, document real CAC, LTV, repeat rate, on-time delivery, dispute rate. Turns the conservative financial model from a projection into evidence.
Founding engineer addresses technical key-person risk. Advisory board addresses strategic + network key-person risk. Recruit and announce 3–5 named advisors from marketplace ops, AI/dev-tools, legal/finance.
2 of the original 4 asks were resolved in v6: "Pick a primary market" (committed EU-first) and "Add conservative-case financials" (added to plan). Remaining 2 are execution milestones that require time, not just planning.
All 3 red flags resolved. Conservative case still produces a $9.5M Year 3 outcome. Structural advantages (EU network + junior supply) are genuine. Plan shows discipline at every stage. Invest $250K for 5% post-money ($5M). Revisit at seed after first 20 projects confirm the unit economics.
The AI plays a seed-stage investor at a respected fund. System prompt directs: challenge every claim, quantify every risk, be harsh but fair, never cheerlead, never invent data, reference concrete sections of the plan.
Team, Product, Market, GTM, Finance, Defensibility, Execution, Legal, Exit. Each dimension yields red/yellow/green flags. Overall verdict is weighted toward the two most critical: Team and Traction.
Every plan revision is analyzed. Flags migrate: red can resolve to yellow or green; green can degrade if new information emerges. Downgrades are called out explicitly — no hiding.
Structured JSON: verdict, confidence, flags (by color), pending asks, iteration diff. Both human-readable on this page and machine-readable for the platform's live Investor Lens on generated specs.
Why publish this? An investor reviewing Suprance would run this exact analysis privately. We run it publicly — in front of the investor, before they ask. That's either extremely dumb or extremely honest. We're betting on honest. Investors who value adversarial self-review will find us more credible, not less.