Paste your startup idea. Our AI plays a sharp angel investor who’s seen 500 pitches — specific, sometimes savage, genuinely trying to help you win.
Use these free tools to take the next step.
The overall score is a weighted average of six dimensions. Here is how to read the number — and what to do with it.
Rare. Strong signal across most dimensions. Worth pursuing aggressively.
Good bones. One or two weak areas that can be addressed before you build.
Real problems in the idea. Not unfixable, but don't start building yet.
Fundamental issues with the premise. Consider a significant pivot before proceeding.
The AI found no viable path forward with the idea as described. That's useful information.
Each dimension is scored 1–10 and weighted differently in the overall score. Understanding what each one measures helps you target the right improvements.
Is the pain real, specific, and urgent enough that someone would pay to fix it? Vague problems score low. Sharp, documented pain scores high.
Is the addressable market large enough to build a real business? A niche within a niche scores low. A $1B+ obvious opportunity scores high.
How defensible is the space? A bloodbath with entrenched players scores low. A clear gap with no dominant player scores high — counter-intuitively, high competition score = good.
Is there a clear, plausible path to revenue? "We'll figure out monetisation later" scores low. Obvious unit economics score high.
Is now the right moment — not too early, not too late? A wave of regulation, tech, or behaviour change creating the window scores high. Already commoditised scores low.
Can you tell exactly who pays, how much, and why from the description alone? Fuzzy buyer and willingness-to-pay scores low. Crystal-clear pricing and buyer score high.
The Roast My Idea tool is designed for founders who want an honest answer, not a comfortable one.
You have an idea you're excited about but you're not sure if you're being realistic. Use this before you spend months building.
You want an honest outside view before you commit your evenings and weekends. 30 seconds to find out if it's worth the sacrifice.
You've built before and you know the mistakes. Use this to find the blind spots in your current idea before a VC does.
If you want someone to tell you your idea is great, this is the wrong tool. The AI is prompted to find what's wrong, not what's right.
These are the patterns the AI looks for and penalises heavily — because investors do too.
If there are genuinely no competitors, there is probably no market. Real markets have competition. The question is whether you can win in a crowded space or whether there's a defensible niche competitors are ignoring.
Top-down market sizing tells investors nothing. What they want to know is: who is your first 100 customers, and why will they switch? Bottom-up beats top-down every time.
1% of a large market sounds conservative but it implies enormous distribution reach. Early-stage startups need a beachhead — a specific customer segment they can dominate completely before expanding.
No product sells itself. Not identifying a clear customer acquisition channel is the #1 reason well-built products die. The roast tool specifically evaluates whether your business model implies a workable go-to-market.
The roast tells you what's broken. Here is what to do about each type of score.
Talk to 10 potential customers before you do anything else. Not surveys — actual 20-minute conversations. Ask what they spend money on today to solve this problem. If they don't spend anything, that's your answer.
Either the market is genuinely small (reconsider), or you described it too narrowly. Try reframing the problem one level up. A tool for dentists can become a tool for all healthcare practitioners — if the problem is actually the same.
Find a specific beachhead — a subset of the market that incumbents are underserving. Compete in the niche first. You don't beat Salesforce by being a better CRM; you beat them by being the CRM for independent physiotherapists.
Use the Lean Canvas tool to map out your full business model — channels, cost structure, and revenue streams in nine blocks. Then use the TAM SAM SOM calculator to stress-test the numbers.
You have something worth pursuing. The next step is a validation survey with real potential customers and a landing page to test real demand. Fonda's full platform runs this end-to-end.
Each one addresses a specific weakness the roast might have surfaced.
Score above 65?
Fonda's AI co-founder runs the full validation — real customer conversations, demand testing, business plan, and launch strategy — so you know it works before you build.