The goal is not an answer. It is an experiment you can actually run.
“Validate your idea” is advice everyone gives and nobody makes specific. What does the experiment look like? Who do you run it with? What counts as evidence? What would make you stop?
The workspace produces an experiment design — not a score, not feedback, not encouragement. A specific action you can take in the real world to learn whether the core assumption holds.
An idea cannot be validated inside a conversation. It can only be validated in the market.
The workspace does not pretend to know whether your idea will work. No AI can predict that. What it can do is help you identify the one assumption that matters most and design the cheapest possible way to test it.
The experiment is the output. The workspace is the tool that helps you define it clearly.
Action. Audience. Offer. Signal. Decision.
A useful startup experiment has five components:
What you will do in the real world (not what you will think about).
The specific people you will reach, defined narrowly enough to actually contact.
What you will ask them to do or respond to.
The observable behavior that would count as evidence for the assumption.
The observable behavior that would count as evidence against it.
What you will do differently depending on the outcome.
The workspace helps you define each one, drawn from what the other four lenses have surfaced about the problem, customer, demand, and wedge.
The experiment targets the assumption the thesis depends on most.
Not every assumption matters equally. Some are well-supported. Others are completely untested and the entire thesis depends on them.
The workspace identifies which assumptions are doing the most load-bearing work and have the least evidence. The experiment is designed around those — so the founder tests the thing that would change their mind, not the thing that is easiest to confirm.
The experiment runs in the real world. The workspace helps you read the result.
The workspace does not run your experiment. You contact real people, make real offers, and observe real behavior. Then you bring the evidence back.
The workspace helps you classify the result — was this evidence, or another assumption? Did the success signal appear? Should the thesis change?
The conversation continues. The thesis gets stronger or it gets revised. Both are progress.
Questions founders ask first
The workspace helps you design the cheapest real-world action that could produce evidence. This might be a conversation with a potential customer, a landing page with a specific offer, a manual version of the service, or a concierge test. The format depends on what assumption needs testing.
Define the experiment. Then run it.
The workspace takes your idea through five lenses, surfaces the weakest assumptions, and helps you design the cheapest experiment that could change your mind. What happens next is up to you.
Your idea. Your experiment. Evidence from the real world, not from a model.