You validate a business idea by testing it against evidence you don't control for — not your own conviction, not friends who want to be supportive. Describe the idea, run it against real people or real AI agents under acceptance criteria you set in advance, and read the verdict: VALIDATED, INVALIDATED, or UNCERTAIN.
The feedback that doesn't count
Ask your mum if she'd use your app and she'll say yes. Ask a friend at a barbecue and they'll be encouraging. Neither has to live with the consequences of their answer, so neither is a real signal. The opinions worth worrying about come from people with nothing to gain from being nice to you.
What a real experiment looks like
Describe the idea in plain language — no deck, no prototype required. Choose your panel: Human (real verified people), Agentic (AI agents), or Hybrid (both). AI structures the idea into a targeted experiment. You set the budget and define what "validated" means before any data is collected. The experiment runs, and you get a verdict in about 24 hours.
Why "would you use this" is the wrong question
Hypothetical questions get hypothetical, polite answers. A structured experiment replaces the question with a task: real people responding to a real scenario, or real AI agents making a real inclusion/ranking decision. Both produce behaviour, not opinions about behaviour.
What you walk away with
- A VALIDATED / INVALIDATED / UNCERTAIN verdict, with an AI report explaining why
- An Agent-Readiness grade (A–F), if you ran an Agentic panel
- A Smart Validation Certificate, if the idea validated
The idea is free. Knowing whether it's worth building isn't — but it's a lot cheaper than building it and finding out.