5 carousel copywriting frameworks, scripted slide by slide

Why carousel copy needs a framework

A carousel is eight separate decisions to keep going. A single image gets one glance; a carousel earns its reach one swipe at a time, and every swipe is the reader asking "why does the next slide exist?" A framework is the answer to that question, written in advance. Without one, slide 4 follows slide 3 because you had more to say. With one, slide 4 follows slide 3 because the argument needed it there.

This is a different tool from a format. Our list of Instagram carousel ideas covers formats: checklist, myth vs fact, teardown, the shape of the post. A copywriting framework is the persuasion order inside any of those shapes. You pick one of each. The mechanics that reward the combination (dwell, completion, re-serve) are in our Instagram carousel best practices guide; this post is about the words.

One first-hand note. Every carousel generated in xcarousel exists as copy before it is laid out, so we read a lot of carousel text. The failure we see most often is not weak writing on slide 6. It is a slide 1 that labels the topic ("Email marketing tips") instead of opening a loop. The frameworks below fix the middle; the hook section fixes slide 1.

The five frameworks, slide by slide

Every script assumes eight slides, the length viewers finish (the spined listicle takes nine, because its recap slide is non-negotiable). To stretch one, add body slides in the middle. Never split the hook across two slides and never add a second CTA.

1. PAS: problem, agitate, solve

The default for how-to content and anyone selling a service. It works because the reader recognises themselves on slide 1 and wants the exit.

• Slide 1, hook: the problem in the reader's own words. "Your carousels get likes and zero saves."
• Slide 2, agitate: what the problem actually costs. One consequence, one number if you have it.
• Slide 3, agitate: the fix they have already tried and why it did not work.
• Slides 4–6, solve: the solution in three moves, one per slide, each with the action and the expected result.
• Slide 7, proof: a before/after, a screenshot described in words, or the number that changed.
• Slide 8, CTA: one verb. "Save this before your next post."

Watch-out: two agitation slides is the limit. A third reads as manipulation, and the reader who feels manipulated on slide 4 does not reach slide 8.

2. AIDA: attention, interest, desire, action

The cleanest framework for a carousel that sells: an offer, a lead magnet, a launch. Weak for pure education, because desire slides with nothing to want go flat.

• Slide 1, attention: the claim or the number. "We cut carousel production from 3 hours to 12 minutes."
• Slides 2–3, interest: why this matters now, then the mechanism behind the claim.
• Slides 4–6, desire: what it looks like when it works. Specifics only: the workflow, the output, the thing the reader gets to stop doing.
• Slide 7, objection: the "even if" slide. "Even if you have never opened a design tool."
• Slide 8, action: the ask, with the next step named. "Comment CAROUSEL and I'll send the template."

Watch-out: interest and desire blur into each other when every slide is a benefit. Slides 2–3 explain; slides 4–6 show. Keep the explaining out of the showing.

3. BAB: before, after, bridge

Built for transformations, case studies, and tool switches. The reader buys the version of themselves on the "after" slides; the bridge is what makes the after believable.

• Slide 1, hook: the delta as a number. "From 40 hours a month to 4."
• Slides 2–3, before: the old state, honestly, one specific frustration per slide.
• Slides 4–5, after: the new state, the same specifics flipped.
• Slides 6–7, bridge: what changed, one decision per slide. Not the montage, the decisions.
• Slide 8, CTA: the shortcut to the bridge, or "save this for when you are ready to switch."

Watch-out: an after without a bridge is a brag. If you cannot fill slides 6–7 with decisions the reader could copy, this is a story arc, not a BAB.

4. The spined listicle

"7 tips" carousels fail because they are topic dumps: seven unrelated slides with a number on each. The spine is the rule all seven items share, stated once, before the items.

• Slide 1, hook: the count plus the stake. "7 hooks I tested. Only 2 got saves."
• Slide 2, spine: the rule behind the list. "Every hook that worked promised a number."
• Slides 3–7, items: one per slide. The item, why it works (one line), how to apply it (one line).
• Slide 8, recap: all items on one screen. This is the save slide.
• Slide 9, CTA: one verb.

Watch-out: this is the one script that runs nine slides, because the recap is non-negotiable. Skip the spine and you are back to the topic dump this framework exists to prevent.

5. The story arc

The strongest framework for being remembered as a person rather than an account. Use it sparingly and only for a story with a real turn.

• Slide 1, hook: the ending, out of context. "We deleted 40 posts. Reach went up."
• Slide 2, setup: where you were, in one sentence.
• Slides 3–4, struggle: what you tried, what failed.
• Slide 5, turn: the one thing that changed.
• Slides 6–7, after: what happened next, in numbers or specifics.
• Slide 8, lesson: the transferable rule, then one ask.

Watch-out: the turn is a single slide. Stories with three turns are three carousels. (The same arc is one of the six patterns in our breakdown of LinkedIn carousel examples, where the PDF format rewards it even more.)

Carousel hook ideas: 10 formulas that open a loop

A hook is a promise plus a gap. The promise says what the reader gets; the gap makes not swiping feel like a loss. Each formula below is a fill-in-the-blank with one worked example.

1. Number + outcome. "[N] [things] that [measurable result]." → "3 caption changes that doubled our saves."
2. The contrarian claim. "[Common advice] is why [bad result]." → "Posting daily is why your reach is flat."
3. The mistake frame. "You're [doing X] wrong." → "You're writing slide 1 wrong."
4. The ending first. State the result before the story. → "We deleted 40 posts. Reach went up."
5. The cost of not knowing. "This [small thing] is costing you [big thing]." → "This one export setting is costing you the grid."
6. The argument they already have. Name the debate. → "Carousels or Reels in 2026? The data, not the vibes."
7. I did it so you don't have to. "[I tested N things]. [Small number] worked." → "I tested 12 carousel apps. Two are worth installing."
8. The checklist promise. "The [N] checks before you [action]." → "The 6 checks before you post a carousel."
9. The insider reveal. "What [experience] taught me about [topic]." → "What 200 client carousels taught me about slide 1."
10. The specific how. "How a [small thing] beat a [big thing]." → "How a 4-slide carousel outperformed my 10-slide one."

Two rules hold across all ten. Numbers beat adjectives: "doubled" beats "boosted", "12 minutes" beats "fast". And the promise must be one the middle slides actually deliver, which is why the hook is written last.

Instagram carousel hooks that fail

The patterns that look like hooks and do not work:

• The label. "Content tips." "Q3 update." A category is not a promise. Nobody swipes to see a topic.
• The essay. A 25-word hook set at hook size wraps to four lines and reads as a paragraph. If it does not fit in 12 words, the idea is not sharp yet.
• The unbacked cheque. "The secret nobody tells you" followed by advice everyone tells you. The swipe-through rate on slide 2 is where this one dies, and the reader remembers.
• The intro slide. Your logo, your name, the word "Introduction". That is a cover, and covers are not hooks.
• The caption-dependent hook. "Here's what happened next" only works if the reader read the caption first. In feed, most do not. The slide has to stand alone, because in your profile grid it does.

Read slide 1 on a phone at thumbnail size before you publish. If you have to zoom, the viewer will not.

Word budgets per slide

Copy that fits is copy that gets read. These budgets come from what stays legible on a 1080×1350 slide at the text sizes a phone feed needs:

The budgets are also a diagnostic. A body slide that wants 60 words is two ideas fighting for one slide; split it. A hook that needs 20 words is not yet a hook; cut until the promise is the only thing left.

Where the copy goes: slides vs caption

The slides carry the argument. The caption carries everything the argument did not need: context, the link reference, a question that opens the comment thread. Do not repeat the slides in the caption; the reader who finished the carousel gains nothing, and the reader who did not will not start there.

The split matters because most carousel copy already exists somewhere else, as a blog post, a newsletter, a talk. The framework's job is to re-sequence it, not to invent it. If you have the argument written, turning text into a carousel is a splitting problem: the paragraphs become body slides, the strongest claim becomes the hook, and the caption gets the leftovers.

Skip the blank page

Everything above is doable by hand: pick the framework, write eight slides to budget, rewrite the hook last, then design eight slides that hold one visual system and export them at the right size. The writing takes twenty minutes once the framework is chosen. The design takes the hour.

xcarousel collapses both. Name the framework in the prompt ("PAS carousel about why carousels get likes and no saves") together with the topic, a URL, or a pasted draft, and the AI carousel maker writes the slides into that arc, applies a palette and type system, and lays out every slide on an editable canvas. Rewrite the hook until it is yours, export at 1080×1350, post. The framework thinking stays yours; the blank page and the layout hour do not.