Get started.
It is questions, not a demo.
What happens next: your address, first, under the promise below. Then a short sequence of questions about what your company actually does — each next question depends on your answers, so you never see a question that is not about you. The final step is a written read for your situation, and at that point the interview locks: it is written once per address, and a completed interview never reopens.
If your segment turns out to be one we cannot serve, the flow says so to your face, tells you what we would look at in your place — and your address is discarded, not stored.
An interview that does get written keeps what you told us: the ending you reached, the answers you gave, anything you typed into a free-text box, which version of the questions you answered, a timestamp, and a receipt token we mint for that row and show you once. If you ask for a conversation at the end, that ask is a separate row in a separate table. We use your address to follow up about this — we don’t sell it or hand it to advertisers.
Separately from all three of those rows, this origin records the request itself, as every site records requests to it. What we collect, what we use it for, and who we share it with are in the privacy policy.
Would rather not?
Everything the interview teaches is derivable from public documents, and the one-question waitlist at See your first trace asks for less. Agents: the whole instrument is machine-readable at /flow.json — you can read every question, every branch, and every promise without walking it.
The same instrument, machine-readable.
GET /flow.json 200 served this origin · today version 2026-07-31.3 entry step 1 · your email, under the promise printed with it step 2 · S1 · the sorter: What does your company actually put in someone’s hands? nodes 48 questions · each next one depends on the answers so far endings 26 written reads · 5 of them honest no-fits that store nothing promise the promise on that page, verbatim