How experiments work
Each experiment chooses a panel. Human, agentic, or both — depending on who actually buys and recommends your product.
Matched validators from our pool (+35.000 people) answer your experiment. You get emotional signal, willingness to pay, and honest reactions from your actual target.
AI agents will be the next generation of internet users. Today, most products aren't built with them in mind. We help you understand whether an agent would recommend yours — or skip it.
Both panels run in parallel. You see where humans and machines agree - and where they don't. That divergence is the insight no other tool gives you: exactly how much to trust the machine's read.
Tell us about your idea in a quick chat. We'll explain how Validating works and figure out together if it's the right fit for you.
Secure your spot and get priority access to the platform when we launch in Fall 2026 — plus exclusive early member benefits.
Sign up before we launch and €150 in credits are already waiting in your account. Use them on your first experiment the day we go live.
FROM THE BLOG
Ideas, experiments and lessons from the validation lab — for founders who build before they guess.
"Would you pay for this?" gets a yes almost every time. Here's the question that actually predicts whether a stranger hands over money.
CB Insights says 42% of startups die from no market need. AI scoring and emotion detection don't fix it. Here's what actually does.
Alpha testers loved it. Launch day didn't. Here's why validating with people who already like you tells you almost nothing about the market.
Validating is a platform that helps you find out if a business idea actually has demand — before you spend months and money building it. You describe your idea in plain language, we design a short experiment around it, and real people (or AI systems, or both) react to it. You get back a clear report telling you what's working, what isn't, and why — not just a gut feeling.
Here's the shift already underway: more and more of your future customers won't be humans scrolling your website — they'll be AI agents doing the browsing, comparing, and deciding for them. Someone asks ChatGPT "what's the best tool for X" and gets a recommendation instead of a search results page. A shopping assistant compares five products and picks one on someone's behalf. That's not science fiction — it's already happening in a small way, and it's growing fast.
Agentic validation tests exactly that: instead of asking real people what they think of your idea, we give AI systems (the kind behind tools like ChatGPT or Perplexity) a real decision to make — "which of these products would you recommend for this situation?" — and we observe what they choose, and why. We're not asking the AI to pretend to be a human. We're testing the actual judgment of the systems that are increasingly standing between you and your customer. If an AI agent skips your product today, that's a real, usable signal — the same way a human saying "I wouldn't buy this" would be.
Real people — validators from our pool react to your idea. You get emotional signal, willingness to pay, honest human reactions.
AI agents — AI systems evaluate your idea the way they would if a real user asked them for a recommendation. You get a sense of whether tools like ChatGPT would recommend, compare, or skip you.
Hybrid — both, side by side, so you see where humans and machines agree, and where they don't. That gap is often the most useful insight you'll get.
Most founders start with a human panel for their first experiment, then add an AI panel once they want a speed check or a sense of "machine-readiness." Nothing stops you from running all three from day one — it's your call, per experiment, and your waitlist credits work on any of them.
Short, structured tests built around the one assumption that would break your idea if you're wrong about it — never a generic 20-question survey. You describe your idea in plain language, our AI designs the experiment, you review and approve it before anything goes live, and it runs against real people, AI agents, or both. Human-panel experiments typically take a few days; AI-panel experiments can come back in minutes.
No. You can validate an idea that's only in your head, one you've sketched on paper, one you're about to build, or one you already built and something's off. The earlier you validate, the more time and money you save — but it works at any stage.
We recommend a budget per experiment, calculated automatically based on your audience size and the panel type you choose. Waitlist members get €150 in free credits at launch — enough to cover a first experiment, on any panel (real people, AI agents, or both), until the credit runs out. No credit card required to join, no subscription, no automatic charges.
Real, identity-verified people — professionals, consumers, domain experts, freelancers, students — matched to your experiment based on their actual background. Testing a B2B tool for finance teams? You get people with finance backgrounds. Testing a consumer product? You get people who actually buy in that category. They're paid to participate, which is exactly what keeps feedback honest instead of polite.
Yes. Validators and AI agents only see what's necessary to react to your experiment — the problem first, then your solution once the experiment moves into the validation phase. We don't publish your idea anywhere, and you approve every experiment before it goes live.
Fall 2026. Everyone on the waiting list gets priority access, plus €150 in free credits the day we go live.
Validation is the starting point, not the finish line. The moment your results are in, you get a personalised "what to do next" recommendation based on your specific data — not generic advice. From there:
Turn interested validators into first customers. If people showed real interest, you can reconnect with the ones who liked your idea most — a pre-order offer, a beta invite, a founding-member deal. It's all sent through the platform, never a cold outreach; you're talking to people who already told you they were interested.
Build. A landing page pre-filled with your validated messaging, plus discounts on the tools you'll actually need next (payments, email, no-code builders, and more).
Grow. Mentors for things like go-to-market strategy, pricing, or fundraising readiness, plus courses, events, and a community of other founders further down the same road.
Fund. When you're ready to raise, warm introductions to investors actively looking for validated ideas, non-dilutive funding options like grants, and tools to prep your pitch.
Some of this is free, some is part of Ideator Pro — but the point is you're never just handed a PDF and left to figure out the rest on your own.