Give every pre-seed founder the cited research, deterministic financial model, and honest go/hold/no-go call that today only a $1,500–$50,000 consultant provides — in minutes, for a fraction of the price.
The only work item that should start immediately is the validation plan — three tests, $9,500, 45 days. If the demand test and the willingness-to-pay test both fail, this PRD should not be built. If only the reliability test fails, restrict any build to a human-reviewed delivery model.
| Current solution | Key gap |
|---|---|
| LivePlan / Upmetrics ($7–40/mo) | Subscription, template-driven forecasts, not a founder-approved, code-computed model |
| Preuve AI / IdeaProof ($21–499 one-time) | Overlapping scope at lower price; no confirmed deterministic engine or full bundle |
| Business-plan consultants ($1,500–$50,000) | Accurate and trusted, but slow (weeks) and unaffordable pre-seed |
| ChatGPT Plus / Claude Pro ($20/mo) | Cheap and fast, but no deterministic engine, no cited-vs-assumed labeling |
| Category | Metric | Target |
|---|---|---|
| Unit economics | LTV:CAC ratio | ≥ 3.30x base / ≥ 1.20x pessimistic |
| Payback | Payback period | 0 months base / ≤ 4 months pessimistic |
| AI reliability | Citation fabrication rate | < 5%, matching the Phase 0 pass bar |
| AI cost | AI gross margin at p90 usage | ≥ 98.7% |
| Growth | Monthly revenue at month 12 | $14,560 base-case projection |
| Financial health | Runway | 4.5 months on $50,000 pre-revenue |
Day 45. All 3 Phase 0 tests report results; go/hold/no-go decision re-run either way.
~+4 weeks after M0 pass. Intake → decision engine runs end-to-end without manual re-typing.
~+8 weeks after M0 pass. Citation-accuracy eval stays under the 5% fabrication bar.
~+10 weeks after M0 pass. First 5 real paid packs delivered and reviewed.