A first test is a live ad against one problem and one audience. It is not a company. It tells you whether to stop, refine, or keep going — and it should take days rather than quarters.
Talk through the idea
Hold the microphone. Say who hurts, what they try today, why that fails, what it costs them, and what you would charge. Microphone writes a problem statement, hypotheses, and market research you can argue with.
Cut the pitch language
Delete anything where the subject is you rather than the customer. What survives is the sentence you would be willing to spend money against. If a reasonable person could not disagree with it, narrow it until they could.
Generate ads and start a campaign
Generate three to five angles that differ by reason to care, not by adjective. Reject anything you would not say out loud. Point them at one narrow audience on one channel, set a fixed budget and a fixed window, and leave the targeting alone while it runs.
Read the result
A clear winner among angles means people recognize that framing. Flat across honest variations means the claim is wrong, not the copy. Messy means it was not really a test yet. The full method is in how to read ad test results.
If the test has signal
Then find people already talking about the problem. That is when you set up a Pi and connect it to Microphone so social media subagents can seed channels.