A decision, not a summary
Proceed, narrow, investigate, or reconsider — with the score, the confidence, and what would move it.
Multi-agent startup research
Find out what would kill your idea — before you build it.
Scout runs seven specialist agents over live web research, then hands you a scored recommendation, the assumptions it could not settle, and the cheapest experiments that would.
Free account · every claim cited · projects saved
Accounting close copilotv2 · 8 sources
Recommendation
Narrow the focus58/100 · 62% confidence
The pain is real and the timing is good, but distribution is the binding constraint. Narrow to firms already on cloud ledgers.
- Market
- 7
- Timing
- 8
- Competition
- 4
- Distribution
- 3
Assumptions
- Accountants will not trust AI with client booksUntested
- Founder outbound can reach CPA firms economicallyTesting
- What stops the incumbent from bundling this?Contradicted
- 8
- live web searches
- 7
- specialist agents
- 6
- scored dimensions
- 0
- uncited claims
Why Scout is different
Most idea validation ends in a document nobody acts on.
A polished report feels like progress. It is not. Scout ends where the work actually starts — with the questions research cannot answer and a plan to answer them.
What you usually get
- Twelve pages of confident prose
- A market size with no source
- A verdict with no conditions attached
- Nothing to do on Monday morning
What Scout gives you
- A recommendation with a confidence level
- Every claim traced to a source you can open
- The exact conditions that would change the answer
- Three experiments you can start this week
How it works
From a single line to a week of work you can defend.
- 01
Describe the idea
One line is enough. Add customer, pricing, or stage and the research sharpens.
- 02
Watch the research run
Eight searches and seven specialists work in parallel, with every query visible.
- 03
Get a recommendation
A scored verdict with confidence — and the conditions that would change it.
- 04
Leave with experiments
Your riskiest assumptions, each paired with a cheap test that has a failure line.
The workspace
Not a document. A place to decide.
Every finished run opens as a canvas: the recommendation, the thesis it rests on, what is still unproven, and how to find out.
The assumptions still open
Risks and investor objections become rows you set a status on as you learn.
Experiments with a failure line
Every test states what success looks like and what would kill it.
Evidence, sorted by direction
What supports the idea, what cuts against it, and what nobody can answer yet.
Every claim cited
Deduplicated sources you can open, and the Markdown report if you want the long form.
Runs that resume
Completed stages are checkpointed, so a failed run picks up instead of starting over.
The specialists
Seven analysts, one verdict.
Each agent owns one dimension, runs its own live research, and defends its own score. When one fails, the rest still report.
- Market Analyst
- Sizes the market, growth, and why-now timing from live evidence.
- Competitor Analyst
- Maps incumbents, pricing, positioning, and the opening to exploit.
- Customer Analyst
- Extracts real pain, switching triggers, and current workarounds.
- GTM Agent
- Designs a first-customer and first-100 acquisition motion.
- VC Partner
- Asks the skeptical questions an investor will ask on day one.
- Moat Agent
- Judges realistic defensibility for the first 12–24 months.
- Experiment Agent
- Turns your biggest risks into fast, cheap validation tests.
- Synthesis
- Reconciles all seven into one scored recommendation — and owns the disagreements.
What Scout will not do
The limits are part of the product.
A tool that tells you what it cannot do is easier to trust with the things it can.
- Talk to your customers for you. Research narrows the question; only people answer it.
- Pretend to be certain. Every recommendation carries a confidence and the evidence behind it.
- Invent a number. If a claim has no source, it does not reach your canvas.
- Change your thesis without confirmation. Scout can recommend a move; the founder owns the decision.
Stop polishing the pitch. Test the assumption.
One line about your idea is enough to start. You will leave with something to run this week.
Stress-test an idea