FREE — NO SIGNUP
Validate Any Startup Idea
One prompt. Paste it into Claude. Get back a full research report:
Who your competitors are and how much they've raised. How big the market is. What your audience actually says online. How to price it. Where to find your first 100 users. Whether the timing is right. What could kill the idea. And a 90-day action plan.
How to Use
1
Copy the prompt below
Click the Copy button next to the prompt. Replace the placeholder with your startup idea (3-5 sentences).
2
Paste into Claude and send
Works with any Claude plan. For best results, use Cowork which runs the 10 research threads in parallel.
3
Get a full research report
3,000-5,000 words: competitor analysis, market sizing, pricing benchmarks, audience insights, risk assessment, and a 90-day action plan.
10 Research Threads
Competitors
Market Sizing
Audience
Pricing
Distribution
Timing
Technical
Risks
Adjacent Opps
Go-to-Market
= uses live web search = analysis only
The Prompt
You are a startup research director at a top-tier VC firm. I need a comprehensive, investor-grade validation report for a startup idea. You will orchestrate this as a parallel research operation.
===== STARTUP IDEA =====
[DESCRIBE YOUR IDEA HERE — the problem, the solution, target audience, and monetization. 3-5 sentences minimum.]
========================
## INSTRUCTIONS
Execute this research in 3 phases:
### PHASE 1: PARSE THE IDEA
Before launching research, extract the structured components of this idea:
- Problem (the specific pain point)
- Solution (the product/service)
- Target audience (be specific — role, company size, demographics)
- Category (SaaS / Marketplace / Consumer / AI / FinTech / etc.)
- 5 search keywords
- Monetization model
- Unique angle vs. obvious approach
### PHASE 2: PARALLEL RESEARCH
Launch these 10 research threads IN PARALLEL as independent sub-agents. Each thread should search the web thoroughly and work independently. Do not run them sequentially — run all 10 simultaneously.
Thread 1 — Competitor Intelligence (use web search)
Find at minimum 8 direct competitors and 5 indirect competitors. For each: name, URL, what they do, funding raised, key strength, key weakness. Also identify 3+ potential big-company threats, 4+ market gaps no one fills, and assess what moat this startup could build.
Thread 2 — Market Sizing (use web search)
Find real analyst reports and industry data. Provide: TAM with source and year, SAM with reasoning, SOM for first 2 years with assumptions, CAGR with source, 5+ market trends with evidence, adjacent expandable markets, and a reality check on whether a startup can actually capture this.
Thread 3 — Audience Deep Dive (use web search)
Search Reddit, Twitter/X, Hacker News, and niche forums. Deliver: a 150+ word primary persona, 5+ real pain quotes with sources, 5+ online communities with sizes, 4+ current solutions and their frustrations, and a willingness-to-pay analysis.
Thread 4 — Pricing Intelligence (use web search)
Visit actual competitor pricing pages. Deliver: a pricing table of 6+ competitors (model, free tier, lowest paid, mid, enterprise), recommended pricing model with justification, 3 designed pricing tiers with psychology explanation, and a Year 1/Year 2 revenue projection.
Thread 5 — Distribution Channels (use web search)
Research how successful companies in this space acquired users. Deliver: 6+ ranked acquisition channels with estimated CAC and time-to-results, a week-by-week plan for first 100 paying users, 5 specific content pieces to create, viral/referral potential assessment, and 5 specific partnership targets.
Thread 6 — Timing Analysis (use web search)
Search for news from the past 12 months affecting this market. Deliver: timing score 0-100 with justification, 4+ tailwinds with dated evidence, 3+ headwinds, specific 2025-2026 catalysts, window assessment (opening/open/closing), and a comparison to a famous startup that succeeded or failed on timing.
Thread 7 — Technical Feasibility (no web search needed)
Deliver: MVP broken into components with weeks per component, recommended tech stack as a table (layer, technology, rationale), 3+ technical risks with severity, scalability breaking points at 100/1K/100K users, AI/data considerations if applicable, and overall complexity rating.
Thread 8 — Risk Assessment (no web search needed)
Think like a pessimistic investor. Deliver: 3+ critical/existential risks with likelihood and mitigation cost, 4+ moderate risks with mitigations, regulatory/legal risks, team skill requirements and hardest hire, and one detailed hidden risk that 90% of founders would miss.
Thread 9 — Adjacent Opportunities (no web search needed)
Deliver: 5+ expansion opportunities ranked by effort and revenue impact, platform play assessment (could this become a platform?), 5 specific partnership targets with value proposition for both sides, and adjacent markets to enter post-PMF.
Thread 10 — Go-to-Market Strategy (no web search needed)
Deliver: pre-launch plan (4 weeks before), launch week plan (where, what, who), post-launch plan (weeks 2-8), marketing channels as a table (channel, specific tactic, expected result, budget), and a content calendar for the first 3 months.
### PHASE 3: SYNTHESIS
After ALL 10 threads complete, synthesize every finding into a final report. Do not summarize lightly — integrate the actual data, company names, numbers, and insights from each thread.
Structure the final report as:
# Deep Research Report: [Idea Name]
## Executive Summary
Verdict: [GO / CAUTION / NO-GO]
| Score | Rating | Rationale |
|-------|--------|-----------|
| Overall Opportunity | /100 | |
| Market Size & Growth | /100 | |
| Technical Feasibility | /100 | |
| Timing | /100 | |
| Competitive Position | /100 | |
| Uniqueness & Defensibility | /100 | |
One-liner: [One sentence summary]
Key insight: [The single most important finding — the thing that changes how you think about this idea]
Opportunity reframe: [The non-obvious way to think about this opportunity]
What makes this a GO: [2-3 conditions that, if true, make this a great bet]
What makes this a NO-GO: [2-3 conditions that, if true, kill this idea]
## 1. Competitive Landscape
[Full competitor analysis with tables from Thread 1]
## 2. Market Opportunity
[Full market sizing with data from Thread 2]
## 3. Target Audience
[Full audience analysis from Thread 3]
## 4. Pricing Strategy
[Full pricing analysis with tiers from Thread 4]
## 5. Distribution & Acquisition
[Full channel analysis and first-100-users plan from Thread 5]
## 6. Market Timing
[Full timing analysis from Thread 6]
## 7. Technical Feasibility
[Full technical assessment from Thread 7]
## 8. Risk Analysis
[Full risk assessment including hidden risk from Thread 8]
## 9. Adjacent Opportunities & Expansion
[Full expansion paths from Thread 9]
## 10. Go-to-Market Plan
[Full GTM strategy from Thread 10]
## Next Steps: 90-Day Action Plan
| # | Action | Timeframe | Priority | Purpose |
|---|--------|-----------|----------|---------|
| 1 | | Week 1 | Critical | |
| 2 | | Week 1 | Critical | |
| 3 | | Week 1-2 | High | |
| 4 | | Week 2-3 | High | |
| 5 | | Week 3-4 | High | |
| 6 | | Month 1-2 | Medium | |
| 7 | | Month 1-2 | Medium | |
| 8 | | Month 2-3 | Medium | |
| 9 | | Month 2-3 | Medium | |
| 10 | | Month 3 | Low | |
---
Save the full report as `deep-research-report.md` in the current folder.
## CRITICAL RULES
- Use web search aggressively in Threads 1-6. Real data only — do not fabricate company names, funding amounts, or market figures.
- Every number needs a source or must be explicitly labeled as an estimate.
- Be brutally honest. A harsh CAUTION is more valuable than an unearned GO.
- The report should be 3,000-5,000 words. Do not compress — this should replace weeks of manual research.
- Run the 10 threads in parallel, not sequentially. This is the entire point.
Tips for Best Results
●
Be specific about your idea. "AI invoicing tool" gives mediocre results. "AI tool that auto-generates invoices from Trello/Asana for freelance designers" gives great results.
●
Cowork runs the 10 threads in parallel. Regular Claude chat runs them sequentially which is slower but still works fine.
●
If Claude truncates the output, send a follow-up: "Continue from where you left off. Do not repeat previous sections."
Already know competitors exist? Search 3.5M+ startups instantly.
Try Brainwave Search