Most ideas die in a document. The expensive ones die after you have already built them. Microphone is for the test in between — the one that costs a weekend instead of a season.
What you do
Pick one problem, one audience, one offer. Talk it through. Generate ad angles and run them against a narrow audience for a fixed window. Decide before launch what each outcome will mean, so that reading the result is analysis rather than narration.
Why narrow beats broad
A broad audience always produces mediocre numbers, and mediocre numbers are unreadable. Narrow until you could name three real people who fit. If the test fails against a specific group, you learned something. If it fails against everyone, you learned nothing.
What a result entitles you to
Conviction, and nothing more — permission to spend the next few weeks. It is not product-market fit and not a reason to skip the hard product work. The distinction matters enough that we wrote an ad is not a company about it.
Start with your first test, or read the full method in how to test a startup idea with ads.