Verified against Claude · 2026-07-23
Outline a swipeable LinkedIn carousel from a single idea
Turn one idea into a slide-by-slide outline for LinkedIn's native document-upload carousel format, built around exactly one point per slide and a design note per slide, so it reads as a designed sequence instead of a wall of bullet points split across pages.
The prompt
Ready to copy — highlighted parts are example details you can swap.
CORE IDEA A 4-step framework for triaging inbound feature requests without a PM AUDIENCE Early-stage founders who are still doing product management themselves ROUGH SLIDE COUNT 8 SPECIFICS THAT MUST APPEAR (data, steps, examples) The 4 triage buckets: fix now, roadmap, decline with reason, needs more data DESIGN CONSTRAINTS Making slides in Canva, text and simple icons only, no custom illustration WHY EVERY SLIDE HAS TO EARN ITS OWN SWIPE This will be uploaded as a PDF through LinkedIn's native document-post feature, which renders as a swipeable carousel directly in the feed. Each swipe is a fresh decision point where a reader can stop — unlike a single post's one-time scroll-past risk, a carousel loses a share of its audience at every slide, so a slide that doesn't clearly earn the next swipe is where the carousel actually dies, even if slide one was strong. STRUCTURE COVER SLIDE (slide 1) — a hook line that works exactly the way a post's first line does: a claim, a number, or a real tension, never a title-card phrasing like "5 Tips for X" or "A Thread on Y." The cover has to make someone want to see slide 2 specifically, not just describe what's coming. BODY SLIDES (slide 2 through second-to-last) — exactly one idea per slide. Each gets a short slide title under 8 words and 1-3 lines of supporting text — no slide should require a reader to actually read a paragraph to get the point; if it does, that's two ideas fighting for one slide, so split it into two. FINAL SLIDE — the single most important takeaway, stated once, plainly, plus a soft call to action: follow, save, or a genuine question — never a generic "Like and share!" disconnected from what the carousel actually delivered. FOR EVERY SLIDE, INCLUDE ONE DESIGN NOTE A short visual direction — "big number as the sole visual anchor," "before/after split down the middle," "one quote, nothing else on the slide" — so this reads as a designed sequence rather than slides of bullet points pasted into a template. If two consecutive slides would use the same design note, vary at least one of them so the sequence doesn't visually flatten. OVERLOAD CHECK After drafting, review every slide against the one-idea rule. Flag any slide that's actually trying to carry two ideas at once — a step plus a caveat, a number plus an unrelated tip — and split it into two slides instead of leaving it dense. CONSTRAINTS FROM SPECIFICS AND DESIGN Every fact, step, or example named in The 4 triage buckets: fix now, roadmap, decline with reason, needs more data must appear on some slide — do not drop one because it doesn't fit the slide count cleanly; adjust the slide count instead and say so. Respect Making slides in Canva, text and simple icons only, no custom illustration on what's actually producible (for example, text and simple icons only, no custom illustration) rather than suggesting a design note that isn't achievable with the stated tools. OUTPUT A slide-by-slide list: slide number, title, supporting text, design note. Close with the overload check results and, if the ideal slide count differs from 8, the adjusted count and why.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Native document posts render as a slide-by-slide swipe, and every swipe is a separate dwell-time event the algorithm can register — this is the documented mechanical reason carousels tend to out-reach a text post of comparable length: the format itself manufactures more total time-on-post than a single scroll past one block of text, but only if a reader keeps swiping. That advantage is conditional, not automatic, which is why the one-idea-per-slide rule and the explicit overload check matter more here than in almost any other format. A carousel that asks a reader to parse a paragraph on slide 4 gives them a natural exit point mid-sequence, and because the entire reach benefit depends on completed swipes, one dense slide can cap the carousel's performance regardless of how strong the surrounding slides are. Requiring a distinct design note per slide, with variation flagged between consecutive slides, targets a specific and common carousel failure: a deck that is functionally a slideshow of identical bullet-point templates with new text dropped in reads as a document, not a designed sequence, and readers can tell the difference within the first two slides — the design note forces an actual visual decision for every slide rather than a template default. The instruction to adjust slide count rather than drop a required fact from {{specifics_to_include}} protects against the failure mode where a fixed slide-count target quietly wins over the content itself. A carousel with the requested number of slides but a missing key data point is worse than one that runs two slides longer than asked but says everything it needed to say.
What you get back
Slide 1 (cover): "Nobody on my team owns product decisions. Here's the 4-bucket system that replaced a PM." Design note: big bold text, no image. Slide 2: "Bucket 1 — Fix now." Text: "Breaks core workflow for >5% of active users. Ships within the week, no discussion." Design note: red accent, single stat callout. Slide 3: "Bucket 2 — Roadmap." Text: "Valid, not urgent. Goes in the next planning cycle, requester gets a real date." Design note: calendar icon motif. Overload check: none of the current 8 slides carry more than one idea — bucket 3 (decline with reason) was originally paired with an example on the same slide, split into slides 5 and 6 instead.
Verified against
Claude Sonnet 5 · 2026-07-23
ChatGPT GPT-5.1 · 2026-07-24
Changelog
- 2026-07-24 — Initial publish for the rewritten linkedin category, verified against Claude (Sonnet 5) and ChatGPT (GPT-5.1).
Need this built into your business?
If a prompt isn't enough — Google Ads management, built and maintained for you — that's Scult's day job.
EXPLORE GOOGLE ADS MANAGEMENT
