The ideation-to-proposal engine: describe your idea, answer the hard questions, and get an investor-ready business proposal that has already survived a five-agent adversarial review.
Entrepreneurs can describe their idea in a sentence, but few can turn it into a business proposal that survives scrutiny. ProposalTank closes that gap with an AI pipeline that writes the proposal, attacks it, and hands back a version that holds up.
Every founder hits the same wall between "I have an idea" and "I have a credible plan." The options today all fail one side of the job. Planning tools like LivePlan and Upmetrics make you fill in a blank template yourself, one field at a time. Idea validators like ValidatorAI and VenturusAI score your idea but hand you a number, not a finished document. One-shot generators like IdeaProof produce a plan but never stress-test it.
ProposalTank does the full loop. A founder describes their idea, works through a structured intake that forces the hard decisions (target customer, revenue model, differentiation), and receives a complete 11-section business proposal. That draft is then fed to a five-agent, cross-vendor adversarial review panel, the same engine that powers VerdictTank, which attacks it for flaws, contradictions, and blind spots. The founder gets back a revised proposal that has already been beaten on.
| Component | Status | Capability |
|---|---|---|
| Proposal generation pipeline | Built and in use | Three-tier AI team ($1.50 / $3.00 / $6.00 per proposal) produces full 11-section proposals. Currently a skill, not a product. |
| Adversarial review engine | Built and live | VerdictTank's 5-agent cross-vendor panel. Live at verdicttank.com, priced $19-$299/mo. |
| Scoping / competitive intel | Built | Phase 1 research workflows: competitor landscape, market sizing, pricing bench. |
| Revision loop | Manual today | Verdict findings are fed back into a rewrite by hand. This is the piece to productize. |
| Guided intake (the moat) | Not built | The structured founder-facing questionnaire that forces the hard decisions before generation. |
| Multi-tenant app + billing | Not built | Auth, usage metering, checkout. Standard Stack Auth/Clerk + Stripe pattern. |
ProposalTank is to business plans what VerdictTank is to proposal review: an AI that does not just write your plan, it beats it up and hands you the version that survives.
Describe your idea in a sentence. Answer the questions the tool forces on you, the ones you would rather skip, the ones that determine whether the business lives or dies. In under an hour you get back an investor-ready business proposal that has already been attacked by five independent AI reviewers and revised around every flaw they found.
Not a blank template. Not a vibes score. A validated proposal, on demand.
The distance between a founder's idea and a credible business proposal is where most ventures die before they start. The tools meant to close it all fail one half of the job.
A founder who wants funding, a co-founder, or a bank loan needs more than a good idea. They need a written proposal that answers the questions an investor actually asks: who is the customer, how do you make money, why can no one copy you, what kills this. Producing that document is hard, slow, and expensive, so founders either skip it, fake it with a template, or pay a consultant thousands of dollars.
| Today's options | What they do | What they leave out |
|---|---|---|
| Planning tools (LivePlan, Upmetrics, IdeaBuddy) | Fill-in-the-blank business plan software with AI assist | You still write it. No one attacks the logic. Output is a plan, not a validated proposal. |
| Idea validators (ValidatorAI, VenturusAI) | Score your idea, run SWOT/PESTEL/Porter's | Hand you a number and a framework. No finished document. No revision loop. |
| One-shot generators (IdeaProof, PrometAI) | Generate a plan from a text prompt | Produce once, never stress-test. Garbage in, garbage out, with no adversarial pass. |
| Human consultants | Write the plan for you | Expensive ($1k-$10k+), slow (weeks), and still one human's opinion. |
| Doing nothing | Wing it | The default. Founders show up to investors with a pitch deck and no defensible plan. |
Who this is for: first-time and solo founders, non-technical founders, side-project builders, and grant/accelerator applicants who need a credible, investor-ready proposal but lack the time, money, or expertise to write one from scratch and have no one to pressure-test it.
Who this is not for: funded startups with in-house strategy teams, and anyone happy to fill in a template themselves. ProposalTank is deliberately scoped to the founder who needs a validated proposal, not a general business-plan editor.
The market is the founder who needs a credible proposal. That is a large, steady pool, and the competitive landscape proves the whitespace is the review loop, not the writing.
| Layer | Estimate | Basis |
|---|---|---|
| TAM | Tens of millions of new ventures / year | ~5M new US business applications / year (Census Bureau Business Formation Statistics); global formation far larger. Assumption, flagged for validation. |
| SAM | ~10-15% of TAM | The subset that needs external validation (funding, co-founders, grants, loans), not lifestyle businesses. Assumption. |
| SOM | Low thousands in year one | Self-serve, English-speaking, online-reachable founders in founder/indie communities. Assumption. |
| Competitor | Price | Category | Does it review the plan adversarially? |
|---|---|---|---|
| LivePlan | $15-$30/mo | Planning tool | No |
| Upmetrics | ~$14/mo | Planning tool | No |
| IdeaBuddy | ~$15/mo | Planning tool | No |
| ValidatorAI | Free | Idea validator | No (score only) |
| VenturusAI | Not captured | Idea validator | No (frameworks only) |
| IdeaProof | From ~EUR0.50 / asset | One-shot generator | Partial (multi-model cross-check, no revision loop) |
| PrometAI | Not captured | One-shot generator | No |
Closest competitor is IdeaProof. It already runs the idea-to-launch motion (validate, generate plan, brand, pitch deck, landing page) at commodity pricing. It validates with multiple models but does not run an adversarial revision loop, does not force the founder through a structured intake, and its output is a plan, not a proposal. It is the player ProposalTank must differentiate against hardest.
ProposalTank is a six-step pipeline, not a text box. The founder moves through it gate by gate, and the product does the heavy lifting at each step.
The output is a real business proposal, structured the way IT Pro Partner already writes them: problem, solution, market, pricing, architecture, deployment options, go-to-market, risks, financials, and the ask. It is hosted at a public URL with the standardized footer, versioned v1 (frozen) and v2 (current), with the pipeline verdict attached. This is the same format and quality bar already applied to VerdictTank, RFP Tank, CartMyList, and Signage.
| Piece | Status | Where it runs today |
|---|---|---|
| Review engine | Live | verdicttank.com, 5-agent cross-vendor panel, $19-$299/mo |
| Proposal hosting | Live | proposals.itpropartner.com (app3) |
| Mockup hosting | Live | mockups.itpropartner.com (app3) |
| Generation pipeline | Built, internal | verdicttank-proposal-pipeline skill (manual orchestration) |
| Guided intake | Not built | To be a founder-facing questionnaire app |
| Multi-tenant app + billing | Not built | Stack Auth/Clerk + Stripe on Core or app3 |
Usage-based, like VerdictTank. The founder pays per validated proposal, not per month of template access.
| Tier | Price | What you get | Cost driver |
|---|---|---|---|
| Spark | $49 one-time | Guided intake + full 11-section proposal draft. No review. | ~$1.50-$6 generation cost |
| Validate | $149 one-time | Draft + VerdictTank 5-agent review + revised v2 + verdict attached. The flagship. | ~$4-$14 generation + review cost |
| Founder | $299/mo | Unlimited proposals and review cycles, RFP Tank compliance scoring, priority queue. | ~$30-$60/mo variable at heavy use |
Two moats, plus a head start no competitor has: the review loop, the intake, and a live review engine already in production.
| Player | Writes the proposal | Attacks it adversarially |
|---|---|---|
| LivePlan / Upmetrics / IdeaBuddy | You write it, they format it | No |
| ValidatorAI / VenturusAI | No | Scores it, but no finished document |
| IdeaProof / PrometAI | Yes, one-shot | No revision loop |
| ProposalTank | Yes, full 11-section | Yes, 5-agent cross-vendor + revision loop |
Generation is a commodity now. Any tool can call an LLM and emit a plan. What none of them does is attack the draft with five independent, cross-vendor reviewers, then revise around every finding and hand back the version that survived. That loop, generation to adversarial review to revision, is a product architecture decision, not a feature, and it is what turns a text generator into a validation engine.
Output quality is bounded by input quality. A free-text box ("I want an app for dogs") produces a generic plan. ProposalTank's structured intake forces the founder to commit to the answers that determine survival before a single token is generated. This is a workflow and prompt-engineering asset that compounds, and it is hard to copy because it encodes IT Pro Partner's actual proposal methodology.
The review engine is not a slide. It is running today at verdicttank.com with a 5-agent cross-vendor panel and paying pricing tiers. ProposalTank embeds it from day one. A competitor would have to build and validate that engine before it could match the loop.
Depth-first, one community at a time. Win the build-in-public crowd first, then move upmarket to accelerators and agencies.
| Channel | Motion | Why |
|---|---|---|
| Build-in-public | Show real proposals it produced, including its own | Trust via transparency; the output is the marketing |
| Product Hunt launch | Launch the "ideation to proposal" story | Founder-heavy audience, one launch spike |
| VerdictTank cross-sell | Every VT review ends with "want the generator?" | Existing audience, zero CAC, the loop sells itself |
| Founder communities | Answer "how do I write a business plan" threads | High-intent, low-cost, repeatable |
Target the first 100 at the Validate ($149) tier through build-in-public trust and the VerdictTank cross-sell, not paid ads. The metric that matters is not signups, it is proposals that pass review and get shared, because a shared proposal is the only marketing that compounds.
The risks are real and mostly about quality and positioning, not technology. Each has a mitigation already on the roadmap.
| Risk | Likelihood | Impact | Mitigation |
|---|---|---|---|
| IdeaProof undercuts on price, starts a commodity race | High | Medium | Never compete on raw text price. Compete on the validated outcome and the review loop. |
| Output quality ceiling: weak intake produces a generic plan, founder churns | Medium | High | The intake is the fix. Enforce completion, push back on thin answers before generating. |
| Hallucinated market data in generated proposals | High | High | Mandatory blind audit step is already in the pipeline, plus citation requirement on every market claim. |
| Expectation mismatch: founder thinks "instant company," gets a proposal | Medium | Medium | Position as "validated proposal, not a company." Scope the deliverable explicitly at checkout. |
| Vendor lock-in or cost volatility in LLM spend | Medium | Medium | Pipeline is already cross-vendor by design. The review panel runs five independent models. |
| Brand confusion with VerdictTank (two "Tank" products) | Low | Medium | Clear family positioning: ProposalTank writes it, VerdictTank judges it, RFP Tank scores it. |
| Review verdict gives false confidence | Low | Medium | Present the verdict as input, not a guarantee. Attach the full verdict so caveats are visible. |
Directional scenarios, not forecasts. The unit economics are clean; the open question is acquisition, which the GTM plan tests before any serious build spend.
| Scenario | Volume | Revenue / mo | What has to be true |
|---|---|---|---|
| Conservative | ~20 | ~$3k | VerdictTank cross-sell only, no viral loop |
| Base | ~100 | ~$15k | Build-in-public plus one strong Product Hunt launch |
| Aggressive | ~500 | ~$75k | A proposal goes viral in a founder community |
What is being asked, where it runs, and what is needed to move.
Run ProposalTank on the existing netcup RS 4000/2000 stack plus the Wasabi S3 backup pipeline, the same infrastructure that already powers IT Pro Partner operations and VerdictTank.
Best when: keeping cost and management overhead low, and shipping fast on proven rails.
Dedicated netcup or Hetzner instances with a dedicated S3 bucket, isolated from the shared ITPP stack.
Best when: a paying enterprise customer demands isolation, or load justifies it later.
Shared responsibility: under both options, IT Pro Partner manages everything below the application layer (compute, network, storage, backup, recovery). The customer owns only their data and usage.
| Name | Domain | Status |
|---|---|---|
| ProposalTank | proposaltank.com | Available |
| IdeaTank | ideatank.io | Available (.com taken) |
| FoundryTank | foundrytank.com | Available |
| Decision | Status | Note |
|---|---|---|
| Product name (ProposalTank vs IdeaTank vs other) | OPEN | proposaltank.com is available; name is a working title |
| Lead tier (Validate $149 is proposed) | OPEN | Confirm price before building checkout |
| Build order: intake first, then multi-tenant app | OPEN | Pre-mortem argues intake first |
| Bundle with VerdictTank or keep separate product | OPEN | Family story: ProposalTank writes, VerdictTank judges |
| Deployment target (Option A shared is proposed) | OPEN | A is default for cost and speed |