Right now
Waiting to start
Your idea, measured.
It looks for people who already want what you are building. If it cannot find them, it says so.
Measure my ideaReplay A real case, 16 Aug 2026. Not your idea. Yours starts below.
Just found
Nothing yet.
So far
0sources confirmed
Verdict
Not until it has read everything.
Every quote is real
Tap one. See the page it came from.
Nothing is paraphrased. The sentence in the case file is the sentence on the page, and the page is one click away.
12 pages were photographed for this case. 8 came back. 4 refused an automated screenshot and are shown with the reason instead of a picture.
How it decides
The verdict is a sum, not an opinion.
26 questions, 130 points, floor at 65. Each question is scored against what it found, or it abstains and says why.
A real one
64 / 130
Don't build · one point under the floorTracking iCal feed latency across O365/Google/Apple
Where it lost: willingness to pay, 3 of 15. No explicit statements of willingness to pay.
Where you land
95 ideas have been through this exact test. Yours goes on the same line.
80 passed the floor, 11 did not. Four scored above 100 and are recorded as 100. If yours is a no, the three nearest that passed light up.
One mark per idea graded on this rubric, from the machine's own records. Four more were graded but their scorecards are missing, so they are not drawn.
Your turn
Describe your idea.
Two sentences is enough. Who has the problem, what it costs them, and what you would build.
Free
The verdict, the score, and three sources you can open.
$49
Every source, the rivals and their prices, and everything it could not confirm.
$79
All of that, plus the build spec written for a coding agent.
Purchases are paused while the desk is rebuilt. The measurement above is free and still runs.
One-time. All sales final. Bunbun Labs Limited, United Kingdom.