Carousel hooks fail for one of four reasons: the first slide doesn't stop the thumb, it doesn't promise anything specific, it doesn't tell the reader what kind of post this is, or nothing on it says there is more. Fix the first three in one short headline, ideally 8 words or fewer, set big enough to read at feed size. Fix the fourth with a visual cue that points to slide 2, and make sure slide 2 still makes sense on its own.
Disclosure: Oakgen is our product. Oakgen Carousels has a cover picker and a hook check built around these rules, and we cover what it does and doesn't do near the end. All hook examples on this page are invented for illustration. They are not real posts and come with no results attached.
Write three hooks for your next carousel
Draft a carousel from your brand in Oakgen Carousels, then compare three alternative covers side by side before you post.
What an Instagram carousel first slide must do
You get one thumb-flick. In that time the first slide has four jobs.
| Job | What it looks like | How it fails |
|---|---|---|
| Stop | High contrast, one large headline, a shape or face the eye lands on | Pale text on a busy photo; three competing elements; small type |
| Promise | A specific outcome, answer, list or story the reader will get by swiping | A topic label ("Productivity tips") instead of a promise |
| Orient | The reader can tell it's a list, a story, a comparison or a fix | A clever line that could be about anything |
| Point onward | Something says there is more: a counter, a peek, a line crossing the edge | A complete-looking poster with nothing pulling right |
Most weak covers do the first job and skip the rest. A striking photo with "Monday thoughts" on it stops people and then lets them go, because nothing tells them what the swipe pays.
"Orient" is the one people forget. A reader decides to swipe partly on the shape of the post. "6 clauses every freelance contract needs" tells you it's a list of six, that it's short, and that you can save it. "Let's talk contracts" tells you nothing, so the reader has to gamble a swipe to find out. Most won't.
8 carousel hook patterns, with examples
These are the shapes most carousel hooks fall into. None is better in general. Pick the one that matches what your slides actually deliver, because a hook that promises a story and delivers a list feels like a trick.
| Pattern | Template | Invented example | Best when the slides are... | Where it breaks |
|---|---|---|---|---|
| Curiosity gap | "The [thing] most [people] skip" | "The oven setting most home bakers skip" | One key insight with steps after it | When the reveal on slide 2 is obvious |
| Specific number | "[N] [specific things] [audience] needs" | "6 clauses every freelance contract needs" | A list, one item per slide | When the number is vague ("some tips") or padded |
| Contrarian | "Stop [common habit]" | "Stop stretching before your easy runs" | An argument with reasons and proof | When you can't back the claim on later slides |
| Before / after | "[Old version] vs [new version]" | "Our first menu vs the one that sells" | A transformation with a visible change | When the "after" is only slightly better |
| Mistake | "The [topic] mistake that costs [audience] most" | "The budgeting mistake that costs renters most" | A problem, its cost and the fix | When the mistake is one nobody makes |
| Mini-story | "[Event]. Here's what changed." | "A client fired us. Here's what changed." | A story arc with a lesson | When the story is about you and not the reader |
| Question | "Is your [thing] [quietly wrong]?" | "Is your emergency fund in the wrong account?" | A diagnosis the reader wants to run | When the answer is yes/no and obvious |
| Checklist | "Before you [action]: [N]-point check" | "Before you list your flat: 9-point check" | A save-worthy reference | When the list is generic enough to find anywhere |
Curiosity gaps need a payoff that's worth it. If slide 2 reveals something the reader half-knew, the next carousel you post gets less benefit of the doubt. Keep the gap narrow and the answer specific.
Numbers work because they set expectations. "6 clauses" tells the reader how long this will take and that each slide carries one item. Use the real count. If you have four good points, the hook says 4, not 7.
Contrarian hooks carry the most risk. They earn attention by disagreeing with something the reader believes. That debt gets paid on slides 2 to 5 with reasons, and on a proof slide if you have one. If all you have is the opinion, the comments will point it out.
Mini-stories should turn toward the reader fast. "A client fired us" is about you. Slide 2 has to make it about them ("Here's the email that started it, and what to check in yours").
Questions are the weakest default. A reader can answer "Do you drink enough water?" without swiping. The question pattern works when the reader can't answer it without your slides: "Is your emergency fund in the wrong account?" makes them wonder which account is the right one.
Carousel hook examples: weak vs strong
Most first drafts are a topic, a feature or a throat-clear. Here are eight, rewritten (all invented).
| Weak hook | Why it's weak | Stronger rewrite | What changed |
|---|---|---|---|
| "Running tips" | A topic, not a promise | "Stop stretching before your easy runs" | A specific habit and a stance |
| "Our new invoicing feature is here!" | About you, not the reader | "Send an invoice in two taps, not ten" | The reader's gain, made concrete |
| "Let's talk about budgeting..." | Throat-clearing; trails off | "The budgeting mistake that costs renters most" | A named mistake and a named audience |
| "5 tips for better phone photos" | Generic list anyone could write | "5 phone camera settings you've never changed" | Specific items the reader probably hasn't seen |
| "Why you should care about your website title and how it can help your business get found online" | 18 words; nobody reads it at feed size | "Your homepage title is costing you clicks" | One claim, 7 words |
| "Wait until you see slide 4" | Hype with no content | "The 4th fix is the one nobody makes" | Same tease, now it promises something |
| "Did you know?" | Answers itself; says nothing | "Your oven runs hotter than the dial says" | A surprising, checkable statement |
| "Swipe to learn more" | Instruction with no reason | "Before you list your flat: 9-point check" | The reason to swipe is the content |
The pattern in the right-hand column: a named reader, a specific noun, and a reason the next slide matters. If your rewrite has none of those, write another.
A good habit is to write ten hooks for every carousel, then cut to three and pick one. The first few you write tend to describe the topic. Keep going until they describe what the reader gets.
How to write a carousel hook that's short enough
Short isn't a style choice on a cover. At feed size on a phone, a long headline either shrinks until it's unreadable or wraps into five lines that nobody reads.
The rule we use in Oakgen's own quality check, and recommend for any tool: warn above 8 words, never go past 12. It's Oakgen's rule, not an Instagram limit, and it's a judgement about reading at feed scale rather than a measured threshold. We also flag covers over about 60 characters, because "Unforgettable" counts as one word but takes up a lot of the slide.
Ways to cut a long hook without losing the promise:
- Delete the setup. "If you're a freelancer, here are the contract clauses you need" becomes "Contract clauses every freelancer needs".
- Move context to a kicker. A small label above the headline ("For first-time landlords") carries the audience, so the headline can carry the promise.
- Push detail to the caption. The caption is where the why-now and the context go. The cover only has to earn the swipe.
- Swap phrases for numbers. "A handful of" becomes "4".
One thing not to cut: the end of the sentence. A hook that ends on "and", "because", a colon or an ellipsis ("Here's why your bread fails because...") is trying to force a swipe by withholding the promise. It looks clever and works against you, because many people never swipe at all, and they've just seen half a sentence. Finish the promise on slide 1. Hold back the payoff, not the promise.
Visual rules for the carousel opening slide
The words do half the work. The other half is whether anyone can read them at the size Instagram shows them.
Contrast first. Dark text on a light field, or the reverse, with nothing busy behind the headline. If you put text on a photo, put it on a calm part of the photo or on a solid band. A headline that's readable on your monitor at full size can disappear on a phone in daylight.
Size for the feed, not the editor. Shrink your cover to roughly the width of a phone screen, hold it at arm's length, and read it. If you squint, the type is too small or too thin. Heavy weights survive compression and small screens better than light ones.
One idea. One headline, one optional kicker, one visual. A cover with a headline, a subheadline, three icons and a logo asks the eye to choose, and it usually chooses to keep scrolling.
Mind the grid crop. Your cover is also your profile-grid thumbnail, which trims the sides, so keep the headline and logo off the left and right edges. The carousel sizes guide has the sizes and crop detail; the Instagram Reels covers guide covers the same problem for Reels.
Design the cover as a cover. The cover sets the visual rules for the rest of the post. If slide 1 is a bold block of brand colour, slides 2 onwards should feel like the same post. The broader principles of stopping a scroll are in our piece on thumb-stopping social media visuals.
Swipe devices: giving people a reason to reach slide 2
A strong hook makes people want the content. A swipe device tells them, visually, that the content continues. These are craft techniques. We know of no public study that measures any of them against each other, so treat them as tools to test.
The peek. Part of a shape from slide 2 pokes in at the right edge of slide 1, so the cover looks unfinished in a deliberate way. The eye reads a cut-off shape as "there's more over there". It costs no words.
The continuous line. One wide canvas sits behind all the slides and is sliced into them, so a line, horizon or path runs from slide 1 across every edge to the last slide. Our working rule here: elements should cross slide edges, one continuous line should run through all slides, and the peek from the right does the pulling. The line can also carry meaning, rising as a list builds or dipping at the problem slide. Keep faces, logos and words off the seams, where they'll be cut in half.
Numbered progress. A small "1 / 7" in the corner tells the reader the post is a sequence and how long it is. It's plain, and that's the point. A reader who sees "1 / 5" knows the commitment is small.
The unresolved sentence, used in the right place. Ending a slide mid-thought pulls people to the next one. Use it between body slides ("The fix isn't the one you'd expect..." on slide 3, the fix on slide 4), never between the cover and slide 2. The cover's promise must be complete for the people who never swipe.
The tease. Slide 1 names something that slide 5 pays off: "The 4th fix is the one nobody makes." The reader now has a reason to get to slide 5, not just slide 2. Only use it if slide 5 actually delivers. A tease with a weak payoff teaches your audience to stop trusting your covers.
What about writing "Swipe" on the cover? The available evidence says it helps, but not much. Socialinsider's 2020 study of nearly 3 million Instagram carousels, reported by Search Engine Journal on 30 August 2020, found that "Telling people to swipe left on a carousel can bring the average engagement rate from 1.83% to 2%." That's an average engagement rate across posts, measured six years ago, and it doesn't separate cause from correlation. A visual cue does the same job without using the words your hook needs.
There's a tension worth knowing about. Continuity devices make slide 2 lean on slide 1. The slide-2 test below asks slide 2 to stand alone. The way through is to let the visual element carry momentum while every slide keeps its own readable headline.
The slide-2 test
Slide 2 isn't only the next slide. In a video posted by Adam Mosseri, the head of Instagram, and reported by Social Media Today on 17 October 2024, he said that if someone sees a carousel but doesn't swipe, Instagram will "often give that carousel a second chance and automatically move to that second piece of media for the viewer". That's a 2024 statement and Instagram may have changed things since, but the practical reading is that some people will meet your carousel at slide 2, cold.
Meta's own Instagram Feed ranking explainer (updated 29 June 2026) also lists "How likely you are to swipe through the whole carousel" among the things it predicts. Our reading: getting people past slide 1 matters, and so does keeping them going after it.
The test takes thirty seconds:
- Cover slide 1 with your hand, or open slide 2 on its own.
- Read it as a stranger who has never seen your cover.
- Ask: does it have its own headline? Does it make sense without the cover? Does it give a reason to keep going?
If slide 2 reads "Here's why:" or "First, some context", it fails. Rewrite it as a second hook that also starts the payoff. For "6 clauses every freelance contract needs", a weak slide 2 is "Clause 1:" followed by body text. A stronger one is "Clause 1: get paid for the work they cancel", which works as an opener on its own.
On LinkedIn, carousels are PDF documents and the opening pages are judged differently. The LinkedIn carousels guide for founders covers that format.
Everything above works in any design tool. The slow part is mocking up several hooks as real covers to see how they wrap at feed size. At one carousel a week that's fine. At four a week, the cover gets rushed first.
Compare three covers on your real slide 1
Oakgen Carousels writes three alternative hooks for your cover and draws each one as your actual first slide, so you can choose by eye.
Testing carousel hooks
You can't A/B test two covers on one Instagram post, so hook testing means comparing posts. Use the 4-week paired test in how to test carousels on your own account, swapping formats for two hook patterns (say, specific number vs contrarian) and holding the layout, swipe device, caption and last-slide ask constant. Write down every hook's exact words and pattern, and compare patterns rather than single posts. Eight posts is a small sample; run it again before you change your style.
Writing carousel hooks in Oakgen Carousels
Oakgen Carousels writes and lays out carousels from your brand book. A few parts of it are built specifically around the first slide.
Pick a cover. In the editor, the "Pick a cover" dialog (its label reads "The first slide decides the swipe") writes three new hooks for your cover, each taking a different angle, and draws each one as your real slide 1, so you see how it wraps and fits before choosing. Each option carries a one-line note on why it might work for your audience. "Three more" writes another set. Only the cover text changes; the rest of the carousel stays as it is. Cover options are free on a paid plan.
The hook check. Oakgen's quality gate has 18 checks, and Approve stays locked while anything blocking is left. "Short, scroll-stopping hook" warns when the cover headline is over 8 words or about 60 characters and blocks it over 12 words. It also warns when the hook runs on to slide 2, such as a cover ending in "because" or an ellipsis.
Slide 2 and swipe cues. "Slide 2 works on its own" warns when slide 2 has no headline of its own. "Each slide hints there is more" warns when a slide before the last has no page counter and nothing crossing its right edge.
One Line layouts and the peek. The One Line series draws one wide canvas behind every slide and slices it, so a line, a horizon and a sun cross the edges. The other three series print a page counter instead. A "Peek at the next slide" toggle adds a shape from the next slide poking in at the right edge. A separate check blocks text that sits on art crossing into the next slide, so nothing important lands on a seam.
The limits, as of October 2026:
- It can't tell you which hook will perform. The three options are drafts with reasons, not predictions. Your own testing decides.
- Hooks stay within your facts. Cover options can only make claims (numbers, results, superlatives) that your approved brand facts back up, which keeps them accurate and sometimes makes them tamer than you'd write yourself.
- Paid plan needed to generate, and artwork costs credits. See pricing.
- Carousel posting goes to Instagram and Facebook Pages only. LinkedIn is a PDF you upload yourself, TikTok is a download, X isn't supported.
When you don't need it: if you post one carousel a week, already have a design template you like, and are willing to run the slide-2 test by hand, the tables on this page will do the job.
Before you post, run the full carousel quality checklist. If you'd rather start from a structure that has already worked for someone else, see how to recreate a carousel that already works. For what the data does and doesn't say about carousels as a format, read do carousels perform better.
Frequently asked questions
What is a carousel hook?
It is the headline on the first slide of a carousel, the line that decides whether someone swipes or scrolls past. A good one makes one specific promise and tells the reader what kind of post they are looking at.
How many words should a carousel hook be?
Aim for about 8 words or fewer, and treat 12 as a hard ceiling. That is a working rule of thumb based on reading at feed size, not a platform limit.
Should my carousel hook be a question?
It can be, but a question is easy to answer in your head and scroll past. If you use one, make it specific enough that the reader cannot answer it without swiping.
Can I split my hook across slide 1 and slide 2?
Avoid it. Many people never reach slide 2, so a promise that only finishes there is one many viewers never see. Finish the hook on slide 1 and let slide 2 start the payoff.
What should go on the second slide of a carousel?
Its own headline that makes sense without slide 1, plus the first piece of what the cover promised. In October 2024, Instagram's head Adam Mosseri said an unswiped carousel may be shown again from its second piece of media, so slide 2 can work as a second cover; that is a 2024 statement and may have changed.
Does writing 'swipe' on the first slide help?
Only a little, on the evidence available. A 2020 Socialinsider study found a small lift in average engagement rate when people were told to swipe; a visual cue such as a shape crossing the edge does the same job without spending words.
How do I test which carousel hook works better?
Change one thing at a time: keep the slides, layout, posting slot and caption the same and vary only the hook pattern across several posts. Judge on one metric chosen in advance, and treat a handful of posts as a hint rather than proof.
Sources
- Adam Mosseri, quoted by Social Media Today, IG chief recommends posting carousels, 17 October 2024: unswiped carousels given "a second chance" from the second piece of media.
- Meta Transparency Center, Instagram Feed ranking, updated 29 June 2026: "How likely you are to swipe through the whole carousel" listed as a prediction.
- Search Engine Journal, Socialinsider Instagram carousel study, 30 August 2020: swipe prompt and average engagement rate, nearly 3 million carousels.
Draft a carousel with a stronger first slide
Build a carousel from your brand, compare three cover hooks, and fix what the hook and slide-2 checks flag before you approve.
