The carousel is the format that most resembles a webpage, which is why it is the easiest one to build from your site and the easiest one to get wrong. A page has sections; a carousel has slides. The danger is treating slides like paragraphs and dumping the whole page across ten frames. A carousel is not a page sliced up. It is a single argument paced across slides, where each slide earns the swipe to the next.
This guide is the slide-by-slide build. For single images see Instagram content from your website ; for video, the Instagram Reels workflow .
Which pages make good carousels
Carousels reward content that has sequence or contrast. A page that is one idea belongs in a single post. A page with steps, a list, a comparison, or a before/after belongs in a carousel.
- A how-to or onboarding doc → “5 steps to X” carousel
- A comparison page → “X vs Y, slide by slide” carousel
- A case study → “How [customer] went from A to B” carousel
- An FAQ cluster → “7 things people ask before buying” carousel
- A mistakes/teardown post → “Stop doing this” carousel
The slide structure
A 7-to-10 slide carousel has named jobs per slide. This is the spine to fill from your page.
| Slide | Job | Rule |
|---|---|---|
| 1 (cover) | Stop the scroll | Big promise, 6 words max, readable at thumbnail size |
| 2 | Earn the swipe | Name the problem or the stakes; confirm slide 1 was honest |
| 3-7 | Deliver | One point per slide, one visual idea per slide |
| 8 | The payoff | The best insight, the slide worth saving |
| 9 | Recap | The whole thing in a list, for re-finding later |
| 10 | CTA | One ask: save, follow, or link in bio |
The cover slide does 80% of the work. It is the only slide that appears in the feed and the grid, so it has to read at the size of a thumbnail. Short, high-contrast, one promise. If the cover says “Some thoughts on content,” nobody swipes. If it says “Write a month of posts in one afternoon,” they do.
Design for the swipe, not the page
Two carousel-specific tactics that have no equivalent on your website:
The cliffhanger edge. End a slide mid-thought so the reader has to swipe to finish it. Slide 4: “The third mistake is the expensive one →”. Slide 5 pays it off. This is the mechanic that lifts completion rate, and completion rate is what Instagram uses to decide whether to push the carousel further.
One idea per frame. Resist putting two points on a slide because they were one paragraph on the page. Crowded slides get swiped past, not read. White space and a single bold line per slide is what gets the whole thing seen.
Worked example: comparison page → carousel
Source: a “vs spreadsheets” page on your site.
- Cover: “Why your spreadsheet is costing you 6 hours a week”
- “You didn’t choose a spreadsheet. You inherited it.” (the stakes)
- Problem 1: version chaos, with a tiny visual of conflicting files
- Problem 2: no history, “who changed this?” →
- Payoff to the cliffhanger: the hidden cost adds up to ~6 hours
- The alternative, shown not sold
- One real number from a customer
- The save-worthy line: “A tool you have to babysit isn’t saving you time”
- Recap: the 3 costs in a list
- CTA: “Save this for your next planning meeting. Link in bio.”
That sequence came straight off one page, but it is paced, not pasted. The same source could also feed a competitor-aware social content plan.
Measure saves and completion
Carousels live and die by two numbers most teams ignore: saves (people keeping it to act on later) and swipe-through / completion rate (how many reached the last slide). A carousel with high saves and 70% completion is working even with modest likes. If completion drops off a cliff at slide 3, slide 3 is boring or the cover over-promised. Fix the slide, not the whole post. Feed these learnings into your social media analytics loop so the next carousel starts from what already held attention.
Utin is being built to draft carousels this way: it reads a page, proposes a cover line and a slide-by-slide outline, flags where the cliffhanger goes, and keeps the source attached for approval. You can register interest for the early pilot.