Verified against ChatGPT · 2026-08-09
Script a carousel slide-by-slide so swipe two through the end earns its place, not just slide one
Turns a single idea into a full slide-by-slide carousel script — hook slide, escalating middle, closer — built around the specific point where most carousels lose their swipe-through and stall.
The prompt
Ready to copy — highlighted parts are example details you can swap.
You are scripting an Instagram carousel, slide by slide, for the idea below. A carousel lives or dies on swipe-through: slide one only has to earn the swipe to slide two, but the real drop-off in most carousels happens around slide three or four, once the initial curiosity from the hook has been spent and nothing new has replaced it.
CORE IDEA
Why batching a week of Instagram content in one sitting produces worse posts than spacing the same work across three shorter sessions.
TARGET SLIDE COUNT
7
AUDIENCE'S STARTING KNOWLEDGE
They've heard 'batch your content' as generic productivity advice and assume more batching is always better.
ONE THING THEY SHOULD DO AFTER
Save the post and try splitting their next batch session into two shorter ones.
BRAND VOICE
Direct and slightly contrarian, short sentences, no motivational-poster language.
PHASE 1 — HOOK SLIDE (slide 1)
Write slide 1 as a single claim, question, or tension statement large enough to read in under two seconds, that makes the specific promise of what this carousel delivers — not a vague teaser ('swipe for tips') but a concrete stake ('the three things about Why batching a week of Instagram content in one sitting produces worse posts than spacing the same work across three shorter sessions. that get people fired from doing this correctly'). State explicitly what slide 1 promises so later slides can be checked against it.
PHASE 2 — MIDDLE SLIDES (build up to slide 7 minus 1)
Write each middle slide as one idea only — never combine two points on one slide, since a slide with two ideas competing for attention reads as neither. Order the middle slides so the information density or stakes increase toward the back half rather than front-loading the best point on slide 2; if the best point is used first, name that explicitly as a risk of what slide would need reordering. Slide 3 or 4 specifically must contain something the audience did not already expect from the hook — a specific number, a named mistake, or a concrete example — since this is the exact point where swipe-through typically drops, and generic restatement of the hook's premise loses people here.
PHASE 3 — CLOSER SLIDE (slide 7)
Write the final slide to explicitly deliver on the promise stated in slide 1 and end with the one post action stated above, worded as a specific, immediate ask rather than a generic 'follow for more'.
WHAT NOT TO DO
Do not write a caption for the post in this pass — carousel slide text only. Do not repeat the hook's exact wording on the closer slide as if restating it were the payoff; the closer should reference the hook's promise while adding the resolution, not echo it.
OUTPUT FORMAT
A numbered list, one entry per slide, each with: the on-slide text (short enough to read at a glance) and a one-line design note (what visual element or emphasis that slide needs). After the slide list, one paragraph flagging where the highest swipe-away risk is in this specific script and what would fix it if testing shows a drop there.Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Requiring slide 1 to state an explicit, checkable promise gives every later phase of the prompt a concrete target to be graded against, which is what turns 'write a carousel' from an open-ended creative request GPT-5.1 will pad with generic advice into a structure it can actually verify slide-by-slide before returning an answer. Naming slide three or four specifically as the highest-risk drop-off point matters because, left unprompted, the model tends to spread its best material evenly or front-load it on slide two out of a generic instinct to 'hook early and often' — but real swipe-through data on carousels consistently shows the mid-carousel slides as where attention actually collapses once initial curiosity is spent, so telling the model where the real risk sits produces a script that spends its strongest concrete detail exactly where it's needed rather than where a generic best-practices template would put it. The one-idea-per-slide rule closes a specific failure mode where GPT-5.1, asked for a fixed slide count on a meaty topic, will compress two points onto one slide to hit the target number rather than cutting content — forcing the count explicit and pairing it with 'never combine two points' makes the model resolve that tension by cutting scope, which is the correct trade-off, instead of silently doubling up a slide. Asking for an explicit swipe-away risk flag at the end forces a self-check pass rather than treating the first draft as final, which is where most of the actual quality gap between a templated carousel and a considered one shows up.
What you get back
Slide 1: "Batching more content in one sitting isn't more efficient — it's slower work in disguise." (Design: bold text, no image yet) ... Slide 4: "By hour three of a batch session, caption quality measurably drops — not because you're tired, but because you've stopped reacting to each post individually." (Design: before/after caption example) ... Slide 7: "Split your next batch into two 90-minute sessions, a day apart. Save this and try it before your next content day."
Verified against
ChatGPT GPT-5.1 · 2026-08-09
Changelog
- 2026-08-09 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
