I built a business idea validator that gives you a straight answer — not just encouragement
After 15+ years managing IT projects, I watched smart teams spend months building things that had obvious structural problems nobody wanted to say out loud. Not bad ideas — ideas with a fatal flaw that a clear-headed 30-minute analysis would have caught.
So I spent a couple of weeks building https://ismyideaanygood.vercel.app — a tool that gives you a structured read on your idea before you commit to it. It's powered by Claude Sonnet, but the lens it applies comes from a ruleset I've built from real project experience: the patterns that actually kill ideas, where founders consistently underestimate costs, and what market signals actually matter.
What you get:
Free snapshot (no login, no email):
- The core opportunity — what actually makes this viable
- The core risk — the one thing most likely to kill it
- A reality check — the uncomfortable thing most validators skip
Full report ($19 one-time):
- Verdict with a score out of 5 and plain-English reasoning
- Market sizing — TAM/SAM/SOM with the caveats most tools skip
- Tech stack + real running costs at 1k and 10k users
- Marketing tactics specific to your idea, not generic advice
- Red flags — the most likely reasons this specific idea fails
- Similar companies — what happened to them and what you can learn
- PDF download to share with co-founders or investors
One feature I'm particularly happy with: the tool detects whether your idea uses AI automatically — no checkbox, no extra step. If it picks up AI in your description, it switches on a cost modelling mode that shows you actual token cost breakdowns at 100, 1k, and 10k users — you can toggle it off if it's not relevant. Most founders building AI products have no idea what their margins look like at scale until it's too late.
There are full example reports on the site so you can see exactly what gets analysed before committing — https://ismyideaanygood.vercel.app/examples
It's been tested across hundreds of ideas and I've reviewed a good chunk of the outputs manually — the quality has been consistently solid. That said, AI can get things wrong. If the tool scores your idea 3 or above, treat that as a prompt to dig deeper and verify the key assumptions yourself, not as a green light to quit your job.
Built with Next.js + Claude Sonnet 4.6. Would love to know: does the analysis feel useful, or does it miss something important for your type of idea?
