The product demo checklist.

Pass one — before capture

1. The outcome is written in one sentence: what the buyer should believe at the end.
2. The audience is named — a role with a problem, not 'prospects'.
3. The stop list exists: six to ten lines of screen → action → what they see.
4. The demo account has plausible names, dates, and numbers. Buyers read the data.
5. One window size for the whole capture; notifications and chat widgets closed.

Pass two — every stop

1. The hotspot sits on the control a real user would click, not on blank space.
2. The tooltip says what the buyer gets ('See which deals stalled'), not what the button is called.
3. One idea per stop. A stop that needs two sentences is two stops.
4. The screen actually changed since the previous stop — no two stops showing the same state.

Pass three — before publish

1. The 'aha' screen — the result — arrives by the midpoint.
2. A branch exists only where two audiences genuinely need different screens.
3. The lead form, if any, sits past the midpoint, after the screen that makes people lean in.
4. The last stop's CTA points at the real next step: sign-up, pricing, or the calendar.
5. You drove the published link in a private window — badge, gate, and all.
6. You drove it on a phone and could reach every hotspot.

Pass four — a week after

1. The per-stop funnel has been read once, seven days in.
2. The stop with the biggest drop got one change — a moved hotspot, a rewritten tooltip, a swapped screen.
3. Any screen the product changed since capture has been re-captured.

Why a checklist and not taste

Every failed demo we have seen fails on one of these, not on aesthetics: a login-screen opening, a form on stop one, a hotspot nobody can find on mobile, a funnel nobody read. The checklist is boring on purpose — it moves the judgement to the two places it belongs, the outcome sentence and the stop list, and makes everything after mechanical.