Use when a long page has all the right sections but they arrive in the wrong order.
page-section-order-argument-flow.md
You are checking the order of the sections on a long page. You are not rewriting the copy inside them.
SECTIONS IN CURRENT ORDER, ONE LINE ON EACH: {{SECTION_ORDER}}
WHAT THE VISITOR ALREADY KNOWS ON ARRIVAL: {{PRIOR_KNOWLEDGE}}
QUESTIONS A BUYER ASKS, ROUGHLY IN THE ORDER THEY ASK THEM: {{BUYER_QUESTIONS}}
SCROLL DEPTH DATA IF AVAILABLE: {{SCROLL_DATA}}
Produce:
1. A table: Position | Section | Question from {{BUYER_QUESTIONS}} it answers | Question it assumes is already answered | Verdict (keep here, move earlier, move later, merge, cut).
2. Every section that answers a question the visitor has not asked yet. This is the main failure mode, so name each one and say which earlier section has to land first.
3. Any question in {{BUYER_QUESTIONS}} that no section answers, and where it belongs.
4. The recommended order in full, with a one line reason for each move you made.
5. Given {{SCROLL_DATA}}, which sections almost nobody reaches, and whether the fix is moving them up or cutting them.
Constraints: work from {{SECTION_ORDER}} as given and do not invent sections. Do not assume a cold visitor if {{PRIOR_KNOWLEDGE}} says otherwise, since the order for a warm visitor differs. Where {{SCROLL_DATA}} is missing, mark point five unavailable rather than guessing depth. No em dashes.
Replace each placeholder with your own detail. The more specific you are, the less the model invents.
When do I reorder rather than rewrite the sections?
When a long page has the right sections arriving in the wrong order and the instinct is to rewrite the copy. Order changes what each section still has to do, so move first and rewrite afterwards. It deliberately leaves the copy inside sections alone, which keeps the argument about sequence rather than about wording.
What do I need in front of me before running it?
One line on each section in its current order, what the visitor already knows on arrival, the buyer questions roughly in the order they get asked, and scroll depth if you have it. Take the question order from sales calls rather than from what the page implies, or you will simply confirm the existing order.
What comes back, and which part is worth acting on?
A position table, the sections answering questions nobody has asked yet, questions no section answers, the full recommended order with a reason for each move, and a scroll depth read. Point two is the main failure mode and the part to act on, since each entry names the earlier section that has to land first.
What is the mistake that costs me here?
Assuming a cold visitor when the traffic is warm. The right order differs, and the prompt follows what you put in prior knowledge rather than a default. Where price should sit is genuinely arguable, since marketing usually wants it later and the visitor wants it sooner, so settle that with scroll and exit data rather than taste.