Test what research cannot settle
Real-world experiments are how a Buteo validation settles what research cannot. For each claim still open after the verdict, Buteo proposes a test, prepares what you need to run it, and weighs what you report back as evidence when the idea is re-evaluated.
By Buteo · Updated
01
The verdict
What research settled, and the claims it could not.
02
Tests are proposed
One per open claim, with what counts as a pass set before you run it.
03
You run them
With the material written for you. It can take days, and the run waits.
04
You report back
What happened, in your words. It is weighed as evidence, like any finding.
The tests it can propose
- Customer interview
- Survey
- Landing page
- Fake door
- Concierge
- Ad test
- Prototype test
- Desk research
What you get
A test per open claim
Chosen for what that claim needs, not a generic checklist.
Pass marks set in advance
What counts as success and what counts as a reason to stop, before any result is in.
Material ready to use
The interview guide, the survey, the page copy or the ad variants, written for your idea.
A run that waits
Take the days a real test needs. The validation is parked, not lost.
Results weighed as evidence
What you report sits in the ledger beside the research, and changes the scores.
A verdict that can move
The idea is re-evaluated with what real people did, not only what was written.
Why run experiments after a validation?
Because research can only read what has been published. It can show that a problem is discussed, that competitors exist and what they charge. It cannot show that the people you would sell to will pay you. Only they can, and an experiment is how you ask them.
So a validation ends with the claims it could not settle and a test for each one, not with a reassuring score.
Which tests does Buteo propose?
The one that fits the open claim, out of 8 kinds: customer interviews, a survey, a landing page, a fake door, a concierge test, an ad test, a prototype test, or more desk research. Each comes with what would count as a pass and what would count as a fail, written down before you run it, so the result cannot be read generously afterwards.
What does Buteo prepare for a test?
The material you need to run it, ready to use rather than described: an interview guide with its questions, a survey with its answer options, the headline and call to action for a landing page, the variants for an ad test. The pass mark is restated in the material itself. You choose which tests to run; none is required.
What happens to my results?
You report what happened and it joins the evidence, weighed beside what research found. The whole ledger is scored again and the idea is re-evaluated, so a verdict can change because of what real people did.
A test can take days, and the run waits for you. Another research pass on the same idea, if you want one, costs 11 credits.
Where it sits on the way to a startup
01 Find the idea
02 Validate
03 Test
- Real-world experiments
04 Build
05 Launch
06 Grow
- Find leadsSoon
- Marketing campaignsSoon
Questions people ask about this
Does Buteo run the experiments for me?
No. Buteo proposes each test, prepares the material for it and weighs the result. Talking to customers, publishing a page or running an ad is yours to do, because the point of a real-world test is that it happens in the real world.
Do I have to run the tests Buteo proposes?
No. The report is the end of a validation and every step after it is optional. You can run some tests, all of them or none, keep validating with another research pass, change direction, or close the idea.